The reception of a foreign head of state, a visiting minister, or a large official delegation is one of the oldest tasks of statecraft, yet it is rarely studied as a subject in its own right. This article treats #diplomatic_hospitality as a distinct field that brings together three connected areas of practice: #protocol, #cross_cultural_communication, and the #logistics of hosting high-level visitors. Using an integrative review of recent scholarship in diplomacy, public diplomacy, intercultural studies, and hospitality management published mainly within the last five years, the paper explains how the way a state welcomes a guest carries political meaning that goes far beyond good manners. It argues that hospitality is a working instrument of #soft_power and #reputational_security, that mistakes in #precedence or #etiquette can damage relationships in ways that are hard to repair, and that the smooth handling of #accommodation, #transport, #security, and #contingency_planning is itself a form of diplomacy. The article also examines the diplomatic table, or #gastrodiplomacy, as a case where culture, communication, and logistics meet in a single event. It closes with practical guidance for #protocol_officers and for students preparing to enter foreign services, tourism boards, and event organizations, and it sets out an agenda for future research. The central claim is simple: hosting is not a background service but a message, and the quality of that message shapes how nations see one another. Keywords: diplomatic hospitality; protocol; cross-cultural communication; logistics; soft power; high-level delegations; public diplomacy; state visits Introduction Every year, thousands of official visits take place around the world. Presidents attend summits, foreign ministers hold bilateral talks, trade delegations tour factories and ports, and royal families make state visits that fill newspapers for days. Behind each of these events is a large team of people who plan seating charts, book hotel floors, arrange motorcades, brief interpreters, taste menus, and rehearse the exact moment when two leaders will shake hands in front of the cameras. This work is what this article calls #diplomatic_hospitality: the organized practice of receiving and caring for foreign guests of state in a way that protects national interests and builds good relations. For a long time, hospitality of this kind was treated as a matter of common sense or good breeding, something that experienced officials simply knew how to do. That view no longer holds. The reception of a #delegation now sits at the crossing point of several serious fields of study. It is part of #diplomacy, because it involves the formal relations between states and is governed by rules such as the Vienna Convention on Diplomatic Relations of 1961. It is part of #public_diplomacy and #nation_branding, because the images produced during a visit reach millions of viewers and shape a country's reputation. It is part of #intercultural_communication, because hosts and guests come from different traditions and read the same gestures in different ways. And it is part of #hospitality_management and #event_management, because a visit is, at bottom, a complex operation involving lodging, food, movement, and time. The stakes are high. A warm and well-run reception can open the door to agreements that had seemed impossible. A cold or clumsy one can sour talks before they begin. History and the daily news are full of small incidents that grew into diplomatic problems: a flag hung upside down, a leader left waiting on a red carpet, a menu that ignored a guest's religious rules, a translated phrase that changed a promise into an insult. None of these mistakes concern the substance of policy, and yet each one can shift the mood of a meeting and, with it, the outcome. This article makes three arguments. First, #diplomatic_hospitality should be understood as a single field with three pillars — protocol, communication, and logistics — that must work together rather than as three separate jobs handled by three separate offices. Second, hospitality is not decoration; it is a channel of power. What a state chooses to serve, whom it seats where, how it moves its guests, and how it handles the unexpected all send signals that other states read closely. Third, because the field is now more visible, faster, and more scrutinized than ever, the people who run it need training that blends the formal knowledge of the diplomat with the operational skill of the event planner and the cultural sensitivity of the communicator. The purpose of the article is both to describe and to guide. For students, it offers a map of a field that is often mentioned but seldom explained. For practitioners, it draws together lessons from recent research that might otherwise stay locked inside separate disciplines. The reader who finishes the article should be able to see why the seating of a dinner, the choice of a gift, and the timing of an airport arrival are not trivial details but decisions with political weight. The article is organized as follows. Section 2 explains the method and scope of the review. Section 3 sets out the concept of #diplomatic_hospitality and its links to #soft_power. Sections 4, 5, and 6 examine the three pillars in turn: protocol and precedence, cross-cultural communication, and the diplomatic table as a special case where culture and logistics meet. Section 7 turns to the wider logistics of hosting delegations, including accommodation, transport, security, technology, and sustainability. Section 8 looks at the people who do this work and how they are trained. Sections 9 and 10 discuss how the pillars fit together and what this means for practice and education. Section 11 notes the limits of the study and points to future research, and Section 12 concludes. Method and scope This article is a narrative and integrative review rather than a report of new field data. Its aim is to gather knowledge that is scattered across several disciplines and to combine it into a single, readable account of #diplomatic_hospitality. This approach suits a subject that has no settled home in the academic world and that is studied, often without being named, by scholars of diplomacy, communication, tourism, security, and management. The review draws on peer-reviewed journal articles and scholarly books, with a clear preference for work published within the last five years so that the picture reflects current practice rather than older habits. Where an older source is a foundation that cannot be avoided, such as the 1961 Vienna Convention on Diplomatic Relations, it is discussed as a primary legal instrument in the text rather than treated as a recent finding. The main bodies of literature consulted were, first, studies of #public_diplomacy, #cultural_diplomacy, and #soft_power, which explain why hosting matters politically; second, studies of #cross_cultural_communication and #cultural_intelligence, which explain how meaning travels between host and guest; third, studies of #gastrodiplomacy and culinary diplomacy, which examine the diplomatic meal; and fourth, studies of #hospitality_management, #event_management, and event #sustainability, which supply the operational side. Sources were chosen for their relevance to the reception and care of high-level visitors, for the quality of the outlet in which they appeared, and for their recency. The review does not claim to be exhaustive. It is selective and interpretive, and it reads the different literatures against one another to find common lessons. Because much of the practical knowledge in this field lives in government manuals and in the memories of experienced officials rather than in academic journals, the review also treats well-documented and widely reported practices as evidence, while being careful to separate established custom from personal opinion. Two limits of the method should be stated at the start. First, a review cannot measure how often hospitality decisions change political outcomes; it can only show that practitioners and scholars believe they matter and can point to patterns that support that belief. Second, the field is uneven across regions and traditions, and the available research leans toward a small number of countries and settings. These limits are revisited in Section 11. What diplomatic hospitality is, and why it carries power At its simplest, #diplomatic_hospitality is the set of arrangements a state makes to receive and look after official guests. But the word "hospitality" hides a long history. In many cultures, the duty to protect and honor a guest is one of the oldest moral rules, older than the modern state itself. Diplomatic protocol grew out of exactly this custom: in early societies, an arriving stranger was offered shelter, food, and safety, and over centuries these acts of welcome hardened into formal rules about how envoys should be received and treated. When a country today rolls out a red carpet, fires a gun salute, or seats a visiting president at the right hand of the host, it is performing a modern version of a very old gesture. The reason this matters for statecraft is that welcome is a language. Joseph Nye's idea of #soft_power — the ability to get what you want through attraction rather than force or payment — helps explain why. A state's soft power rests on its culture, its values, and the way it behaves, and few settings display all three as clearly as the reception of a foreign guest (Nye, 2021). A visit is a stage. The host chooses the set, the script, and the supporting cast, and a global audience watches. Done well, the performance says: we are competent, we are gracious, we respect you, and we are worth dealing with. This is why scholars increasingly treat public diplomacy not only as messaging but as a form of #reputational_security, in which a country's standing is an asset to be protected as carefully as its borders (Cull, 2022). In a tense and media-saturated age, that standing can be helped or hurt in a single afternoon (Cull, 2023). It helps to separate two related ideas. #Cultural_diplomacy is the deliberate use of a country's culture — its art, music, cuisine, and traditions — to build understanding and goodwill with foreign audiences, and it works best when it is designed with clear structures rather than left to chance (Zhu and Wyszomirski, 2023). #Diplomatic_hospitality overlaps with cultural diplomacy but is narrower and more concrete. It is the specific practice of hosting particular people at particular events. When a host serves a national dish at a state dinner, both ideas are in play: the dish is an act of cultural diplomacy, and the dinner is an act of diplomatic hospitality. The political weight of hosting comes from three features. First, it is visible. Unlike closed-door negotiations, arrivals, ceremonies, and banquets are photographed and broadcast, so their symbolism reaches far beyond the room. Second, it is relational. Hospitality is built on the direct, personal contact between host and guest, and the warmth or coldness of that contact colors the relationship that follows. Studies of the small talk that leaders exchange during credential ceremonies show that even brief, seemingly routine conversations are used to build bonds and signal friendliness, which suggests that the "soft" parts of a visit do real diplomatic work. Third, it is reciprocal. Hospitality creates a quiet expectation of return. A state that hosts generously earns a claim on a warm welcome when its own leaders travel, and this exchange of courtesies helps keep relationships steady even when policies disagree. The reciprocal nature of hospitality deserves a closer look, because it is what turns single events into lasting relationships. When one state hosts another with visible care, it creates a debt of courtesy that is expected to be repaid the next time roles are reversed. This exchange runs quietly beneath the formal agenda of a visit. A leader who is received warmly in a foreign capital returns home with a favorable impression that can outlast a particular dispute, and the host, in turn, can expect a comparable welcome when its own delegations travel. Over years, this back-and-forth builds a reservoir of goodwill that steadies a relationship through rough patches. The opposite is also true: a state that hosts poorly, or that appears to downgrade a guest through a cooler reception, sends a signal of distance that the other side will remember. Because these signals are read by professionals who are trained to notice them, the level of hospitality offered on a given occasion becomes a message in itself, quite apart from anything said at the negotiating table. Seen this way, diplomatic hospitality is not a soft add-on to serious diplomacy. It is a working tool that shapes perception, builds relationships, and stores up goodwill. The rest of this article examines how that tool is used, starting with its most formal element: protocol. Protocol and precedence: the grammar of official welcome Protocol is the set of rules that governs how official interactions unfold. It decides who is greeted first, who sits where, how people are addressed, how flags are displayed, and how documents are exchanged. To an outsider it can look like empty ceremony, but its purpose is practical. #Protocol removes doubt. In a setting where several important people from different countries come together, each carrying pride and representing a nation, clear rules prevent the quarrels that would otherwise arise over rank and honor. By fixing the order of things in advance, protocol lets everyone relax and focus on substance. The backbone of modern #protocol is #precedence, the settled order in which people and states are ranked on any given occasion. The Vienna Convention on Diplomatic Relations of 1961 codified much of the older custom on this point. It groups heads of mission into classes — ambassadors and equivalents, envoys and ministers, and chargés d'affaires — and it says that within a class, precedence follows the date and time an envoy formally took up their post. Crucially, the Convention states that, apart from precedence and etiquette, there is to be no difference in treatment between heads of mission by reason of class. In other words, the order is a tool for smooth ceremony, not a ladder of real power. This is why a newly arrived ambassador from a large country may stand behind a long-serving ambassador from a small one: the rule is the date of arrival, not the size of the state. Precedence shapes almost every part of a visit. It decides the order in which delegations enter a hall, the arrangement of flags, the sequence of toasts, and, above all, the seating at meals. Seating is where precedence becomes most delicate. At a formal dinner, the place of honor is traditionally to the right of the host, and the arrangement radiates outward from there in careful steps. Get it right and no one notices; get it wrong and a guest may feel publicly slighted. Because the rules of precedence differ in their details from country to country, and because #cultural sensitivity is now expected, protocol officers often have to blend the strict order with local custom and with the particular sensitivities of the guests in the room. Forms of address and titles form a second strand of protocol. Dignitaries expect to be addressed in a way that matches the office they hold, both in speech and in writing, and errors here are read as ignorance or disrespect. The correct handling of #credentials — the formal letters by which an ambassador is accredited to a host state — is a third strand, and the ceremonies around their presentation are among the oldest surviving rituals of diplomacy. Invitations form a fourth: the wording of a formal invitation, the difference between a formal and an informal event, and the proper way to accept or decline all carry meaning, and a badly worded invitation can confuse or offend before the event has even begun. Symbolism runs through every protocol decision, and the exchange of gifts is one of its clearest forms. A gift given at an official visit is never only an object. It is a message about how the host sees the guest and the relationship between them. A well-chosen gift shows that the host has thought about the guest as an individual and about the guest's country and culture; a careless or generic one suggests the opposite. The best gifts often carry a piece of the host nation's heritage, such as a craft, an artwork, or a product for which the country is known, so that the object itself becomes a small act of cultural display. But gift-giving is also full of traps that differ across traditions. In some cultures certain colors, numbers, or materials carry unlucky or unwelcome meanings; in others, the value of a gift is closely regulated to avoid any appearance of bribery, and an over-generous present can embarrass rather than please. The timing and manner of giving matter too, from whether a gift is opened in front of the giver to who presents it and when. Handling all of this well requires the same blend of cultural knowledge and judgment that runs through the rest of protocol, and it is one more reason the work cannot be reduced to a fixed checklist. Two points deserve emphasis. First, protocol is not frozen. It has changed steadily since the mid-twentieth century to keep pace with new technology, new kinds of actors, and new expectations. Video links now stand in for some face-to-face meetings; multinational companies and non-governmental organizations now hold events that borrow the forms of state protocol; and rising attention to inclusion has softened some of the older, more rigid customs. Second, protocol works best when it is invisible. Its success is measured not by how impressive the ceremony looks but by how completely it prevents friction. When protocol fails, it fails loudly, in a photograph of a snubbed leader or an argument over a seating chart. When it works, the guests simply feel respected and at ease, and the real business can proceed. For students, the lesson is that #protocol is a form of grammar. Just as grammar lets speakers of a language understand one another without stopping to argue about rules, protocol lets representatives of different states interact without stopping to argue about rank. Learning it is less about memorizing a rulebook and more about understanding why the rules exist, so that they can be applied with judgment when situations arise that no rulebook covers. Cross-cultural communication: making meaning travel If protocol is the grammar of official welcome, #cross_cultural_communication is its vocabulary and tone. Hosts and guests bring different assumptions about how respect is shown, how time is used, how directly one should speak, and how much of a message is carried by words rather than by context. When these assumptions clash, even a perfectly organized visit can leave both sides uneasy without either quite knowing why. A useful starting point is the distinction, long used in intercultural studies, between high-context and low-context communication. In high-context settings, much of the meaning lies in the situation, the relationship, and what is left unsaid; in low-context settings, meaning is carried mainly by explicit words. A host who comes from a low-context tradition may value plain speaking and quick agreement, while a guest from a high-context tradition may value the slow building of trust and may read blunt directness as rudeness. Neither is right or wrong, but if the host does not adjust, the guest may feel rushed and disrespected. Related differences in how cultures handle hierarchy and how they balance the individual against the group shape whether a delegation expects visible deference to its leader, prefers group harmony to open debate, and reads silence as agreement or as doubt. The modern answer to these challenges is #cultural_intelligence, often shortened to CQ. Cultural intelligence is the capacity to work well across cultural lines, and researchers describe it as having several parts: the drive to engage other cultures, knowledge about how cultures differ, the awareness to notice and check one's own assumptions in the moment, and the behavioral flexibility to adapt one's actions. In practice, a protocol officer with high cultural intelligence does not simply memorize a list of national customs. Instead, they watch for cues, question their first reactions, and adjust their behavior as a meeting unfolds. Studies of international negotiation find that this blend of self-awareness, cultural knowledge, and adaptable communication helps professionals avoid stereotyping, read situations more accurately, and build the trust on which agreements depend. The same qualities matter just as much at the reception desk and the banquet table as at the negotiating table. There is a warning attached to cultural knowledge, and it is an old one in the field. Categories such as high-context and low-context are useful for training and preparation, but they can harden into stereotypes if applied too rigidly. Cultures are not uniform, and individuals within them vary widely. The skilled host uses cultural knowledge as a first guess to be tested, not as a script to be followed. This is why writers on #intercultural_competence stress that the goal is an open, adaptable disposition rather than a fixed store of facts, and that competence must be renewed as the world changes rather than treated as a box to be ticked (Bo and Onwubuya, 2022). Language sits at the center of #cross_cultural_communication, and #interpretation deserves special attention. In high-level meetings, interpreters are not neutral pipes through which words pass unchanged. They must carry not only the literal meaning but the tone, the level of formality, and the intent behind a phrase. A single mistranslated word can turn a courtesy into an insult or a firm position into a soft one. Good practice therefore treats interpreters as trusted members of the team: they are briefed in advance on the topics and the sensitive points, given glossaries of technical terms, and, where the stakes are high, doubled so that one can check the other. The physical setup matters too — whether interpretation is simultaneous through headsets or consecutive with pauses affects the rhythm of a meeting and must be planned rather than improvised. Two further differences shape how a visit feels to a guest. The first is the handling of time. Some traditions treat a schedule as a firm promise, where lateness is a sign of disrespect and each event should start and end on the dot. Others treat time more loosely, valuing the flow of a conversation or a relationship over the strict clock, so that a meeting that runs long is a sign of engagement rather than a failure of planning. A host who imposes a rigid timetable on a guest from a more flexible tradition, or who keeps a punctual guest waiting, can create friction without realizing the cause. The second difference is the idea of face, or public dignity. In many cultures, preserving the standing of a guest in front of others is a central duty of the host, and anything that risks embarrassing a guest in public, even a minor correction, is avoided. A host who understands this will correct a mistake privately, praise generously in public, and give a guest room to save face when something goes wrong. These are not abstract points; they decide whether a delegation leaves feeling honored or slighted. Nonverbal communication is the quiet partner of language. The length of a handshake, the distance at which two people stand, whether and how they make eye contact, the timing of a bow or a nod, the order in which people are introduced — all of these carry meaning, and all of them differ across traditions. Because so much of a visit is photographed, nonverbal signals are also public. A leader who appears warm and attentive in the images sends one message; one who appears stiff or distracted sends another. This is why the choreography of arrivals and greetings is rehearsed with such care, and why an apparently small lapse, such as a host checking a phone during a welcome, can overshadow hours of careful preparation. The practical lesson is that communication cannot be left to chance or charm. It must be planned as deliberately as the seating chart. Before a visit, the host team should study the guests' communication norms, prepare interpreters thoroughly, brief principals on the nonverbal expectations they will face, and build in ways to check that messages are landing as intended. Cross-cultural communication done well is nearly invisible, because it makes the guests feel understood; done badly, it produces a fog of small misunderstandings that no amount of logistical polish can lift. The diplomatic table: culinary diplomacy as a working example Few settings show the three pillars of #diplomatic_hospitality working together as clearly as the shared meal. A state dinner is at once a ceremony governed by protocol, an act of communication across cultures, and a logistical operation of considerable size. It is also a well-studied example of #gastrodiplomacy and #culinary_diplomacy, which have grown into a lively field of research over the past two decades (Li and Mok, 2025). It is worth separating two ideas that are often confused. #Gastrodiplomacy usually refers to a state's effort to win over foreign publics by promoting its national cuisine abroad, as Thailand did when it encouraged Thai restaurants worldwide and as South Korea, Japan, Malaysia, Taiwan, and others have done since (Li and Mok, 2025). #Culinary_diplomacy, by contrast, refers to the use of food in official, high-level settings — the state banquet, the working lunch, the private dinner between leaders. The parallel is that gastrodiplomacy is to the wider public what culinary diplomacy is to the elite meeting. Both rest on the same insight: eating together is one of the oldest ways humans build trust, and a well-chosen meal can soften a tense relationship in a way that speeches cannot. Food carries diplomatic weight for two reasons that researchers highlight. The first is emotional. Sharing a meal creates a bond and a positive memory, and tasting another country's cuisine is a direct, pleasant encounter with its culture. The second is that food is relatively neutral. Compared with film, literature, or political speech, a dish carries fewer obvious ideological messages, which makes it an easier bridge between countries that disagree on other matters. This is the sense behind the popular description of gastrodiplomacy as winning hearts and minds through the stomach. At the diplomatic table, all three pillars converge. Protocol sets the seating, the order of toasts, and the level of formality. Communication runs through the choice of what to serve and how to present it: a menu can honor the guest by including a nod to their cuisine, or it can showcase the host's own traditions as a form of cultural display. Logistics carries the enormous practical load beneath the surface. The kitchen must produce a large number of covers to a high standard and on an exact schedule tied to the evening's speeches. Every dietary and religious rule among the guests must be known and respected, which requires careful advance work; serving a dish that a guest cannot eat for reasons of faith is not a small slip but a public failure of respect. Allergies must be tracked, security concerns around food handling addressed, and the timing of each course matched to the flow of the event. Recent scholarship on food and diplomacy is expanding beyond the state banquet to include newer actors — diaspora communities, celebrity chefs, and online influencers — who now shape how a nation's cuisine is seen abroad (Octastefani and Kusuma, 2022; Yayusman, Yaumidin, and Mulyasari, 2023). Studies of specific campaigns, from Taiwanese street food and bubble tea to Indonesia's spice-based promotion in new markets, show how governments and citizens together turn food into a tool of image-building (Putri and Baskoro, 2023; Yayusman and Mulyasari, 2024). Reviews of this fast-growing field also caution that much of the research is descriptive and focused on a small number of Asian cases, and that scholars have paid more attention to what senders intend than to what audiences actually feel (Muljabar, 2024; Li and Mok, 2025). For the practitioner, the lesson is that the diplomatic meal is not just a courtesy but a designed message, and that its power depends on how the guest experiences it, not only on what the host meant to say. Logistical management of high-level delegations Beneath every graceful ceremony lies a large and demanding operation. #Logistics is the pillar that the public sees least and that failure exposes most brutally. A missed connection, a broken lift, a security gap, or a hotel double-booking can undo weeks of careful diplomacy in minutes. This section examines the main strands of logistical management for high-level visits: planning and coordination, accommodation, transport and movement, security, technology, sustainability, and contingency planning. 7.1 Planning and coordination A high-level visit is a project with many moving parts and many owners. The host foreign ministry, the visiting delegation's own staff, the security services of both sides, the police, the airport authorities, the hotels, the caterers, the transport providers, and often a private events company all have a role. Coordinating them is the first logistical task, and it is where visits most often go wrong, not because any single team is weak but because the seams between teams are poorly managed. Good practice borrows from #event_management: a single lead coordinator or office holds the master plan, a shared and detailed schedule governs every movement to the minute, and clear lines of responsibility mean that every task has one named owner. The lessons of the wider meetings and events industry — where planners stress that success depends less on the size of the budget and more on defining clear objectives and coordinating many specialist suppliers — apply directly to official visits. 7.2 Accommodation Where and how a delegation is housed sends a message and creates a set of hard requirements. High-level guests expect comfort, privacy, and security, and the standards of #luxury_hospitality — personal attention, a sense of exclusivity, and a smooth, almost invisible service — are exactly what an official visit aims to provide (Christodoulides and colleagues, 2025). Research on what makes a premium guest experience points to the same elements that matter in hosting a delegation: warmth of welcome, anticipation of needs, and consistency across every point of contact. In practice, this means securing appropriate rooms and often whole floors, arranging secure communications within the accommodation, planning private entrances and exits, and coordinating with the hotel on everything from housekeeping schedules to the handling of gifts and documents. The concept of hospitableness — the genuine disposition to make a guest feel cared for — is a useful reminder that the goal is not only physical comfort but the felt sense of being honored. 7.3 Transport and movement Moving a delegation safely and on time is one of the most visible logistical challenges. Airport arrivals are ceremonies in their own right, with their own protocol for who greets whom and in what order. Motorcades must be planned street by street, with routes, timings, and alternatives worked out in advance and coordinated with the police. Even small details, such as which car a principal rides in and who sits beside them, carry protocol meaning. The margin for error is thin: a delegation that arrives late to a ceremony, or that is caught in traffic on the way to a signing, disrupts a schedule that may have been negotiated for weeks. This is why movement is rehearsed, why backup vehicles and routes are standard, and why transport planners work hand in hand with security from the earliest stage. 7.4 Security #Security runs through every part of a high-level visit and shapes decisions that might otherwise be made on comfort or convenience alone. Protecting a foreign principal is a shared duty: the host state carries the main responsibility, while the guest's own protection team works alongside it, and the two must agree on roles in advance to avoid confusion at critical moments. Security concerns influence the choice of venues, the design of routes, the layout of rooms, the vetting of staff, and the handling of food and gifts. The art of good security in this setting is to be thorough without being oppressive; a guest surrounded by an obvious wall of guards may feel protected but not honored. The best arrangements keep the guest safe while preserving the warmth and dignity of the welcome, which requires close cooperation between the security teams and the protocol and hospitality teams so that safety and courtesy pull in the same direction rather than against each other. 7.5 Technology Technology now touches every strand of logistics. Event-management software coordinates schedules, registrations, and communications; secure communication systems protect sensitive conversations; and digital tools help track the many small tasks that make up a visit. The rise of hybrid and virtual formats, accelerated in recent years, means that some meetings now combine people in the room with people on screens, which brings its own protocol and technical demands. Digital diplomacy has also changed how visits are seen: images and messages spread instantly and widely, so the online dimension of a visit must be planned alongside the physical one. At the same time, technology introduces new risks — from cyber intrusion to the rapid spread of an unflattering image — that the modern host must manage. The practical rule is that every technical system used in a visit should be tested in advance and backed up, because a failure in front of the cameras is a failure in public. 7.6 Sustainability High-level visits, like all large events, carry an environmental cost through travel, catering, energy use, and waste, and expectations about #sustainability are rising fast. Research on business and academic events shows growing pressure to reduce this footprint and a growing willingness among participants to support greener practices, though attitudes vary across groups (Santos and colleagues, 2023). Scholars also caution that making events greener is necessary but not sufficient, and that organizers should think about the wider social and economic effects of what they host rather than treating sustainability as a checklist (Mair and Smith, 2021). For the host of an official visit, this means practical steps such as reducing waste at banquets, choosing local and seasonal food, cutting unnecessary travel, and reusing materials where possible — steps that also carry a diplomatic message, since a visibly responsible host displays values that many guests now share. Sustainability, in other words, has itself become part of #nation_branding. 7.7 Budget, stakeholders, and the local community Two practical strands often receive too little attention in accounts of official visits. The first is money. A high-level visit is expensive, and its budget must cover accommodation, transport, security, catering, staffing, venue costs, and the many small expenses that add up quickly. Managing this budget is a diplomatic act in itself, because both waste and visible thrift send messages. A reception that looks lavish beyond reason can draw criticism at home and abroad, while one that looks mean can be read as a slight to the guest. The skill lies in spending where it shows respect and matters to the guest, and saving where it does not, which again requires judgment rather than a formula. The second strand is the wider set of people affected by a visit. Beyond the host and guest teams sit airport authorities, city police, hotel and catering staff, transport firms, and, not least, the ordinary residents of the host city, whose daily lives may be disrupted by road closures and security cordons. Good hosts plan with these groups in mind, communicate disruptions clearly, and try to leave a positive rather than a resentful impression, since the local community is part of the image a visit projects. The meetings and events industry has long understood that an event's success depends on aligning many specialist suppliers and on managing its effect on the host city, and official visits are no exception. 7.8 Contingency planning No plan survives contact with reality unchanged. Weather delays flights, guests fall ill, schedules slip, and unexpected events intrude. #Contingency_planning is the discipline of preparing for these shocks so that they can be absorbed without embarrassment. Good contingency planning identifies the points where a visit is most fragile — the tight connections, the outdoor ceremonies, the single points of failure — and builds in alternatives for each. It also rehearses decision-making, so that when something goes wrong, everyone knows who decides and how the news travels. The measure of a well-run visit is not that nothing goes wrong, since something usually does, but that the guests never see it. A delegation that glides through a day unaware of the problems solved on its behalf has been hosted well. The human factor: protocol officers, staff, and training All the systems described so far depend on people. The #protocol_officer sits at the center of #diplomatic_hospitality, holding together the three pillars of protocol, communication, and logistics. This is a demanding role that calls for an unusual mix of knowledge and temperament: mastery of the formal rules, deep cultural awareness, operational skill, calm under pressure, and the tact to correct a mistake or manage a difficult guest without causing offense. The protocol officer must be able to move in a single day from checking a seating chart against the rules of precedence, to briefing a principal on how a guest's culture reads eye contact, to solving a transport problem, to smoothing over a small crisis at a banquet. Around the protocol officer stands a wider team whose training and disposition shape the guest's experience at every turn. Drivers, greeters, interpreters, catering staff, and security personnel all become, for the duration of a visit, the face of the host nation. Research on service quality in hospitality shows that the guest's overall impression is built from many small moments of contact, and that a single poor interaction can outweigh many good ones. The idea of hospitableness — a real, felt willingness to care for the guest — captures what training should aim for: not merely correct behavior but a genuine warmth that the guest can sense. This is why leading hospitality educators argue that hospitableness must be taught and cultivated rather than assumed, and why the same thinking is now entering the training of official hosts. Education for this field is still uneven. Diplomacy programs teach protocol and international relations but often say little about logistics or service. Hospitality and event-management programs teach operations and guest experience but often say little about diplomacy or #cross_cultural_communication at the level of statecraft. The argument of this article is that #diplomatic_hospitality needs an education that crosses these lines — one that gives the future diplomat the operational grounding of the event planner, gives the future event manager the diplomatic and cultural grounding of the foreign service, and gives both the #cultural_intelligence to adapt in the moment. Practical training methods such as simulations, case studies of real visits, and role-play of difficult moments are especially valuable because so much of the skill in this field is judgment that can only be learned by practice. Discussion: integrating the three pillars The central claim of this article is that #protocol, #cross_cultural_communication, and #logistics are not three separate jobs but three faces of one task, and that the quality of a visit depends on how well they are integrated. The diplomatic table in Section 6 showed this in miniature: a banquet fails if the seating breaks protocol, or if the menu ignores a guest's culture, or if the kitchen cannot deliver on time, and no strength in one pillar can rescue a failure in another. The same is true of a visit as a whole. Integration matters because the pillars constantly interact. A security decision to change a route affects the timing of a ceremony, which is a protocol matter, which affects the images broadcast, which is a communication matter. A cultural insight about a guest's preference for indirect speech affects how a difficult logistical message should be delivered. A sustainability choice to serve local food is at once an operational decision, a cultural statement, and a branding message. When these connections are managed by separate offices that do not talk to one another, the seams show. When they are managed as a whole, the visit feels seamless to the guest, and it is that seamlessness that produces the goodwill the host is seeking. This integrated view also clarifies why diplomatic hospitality is a form of power rather than a service. Each pillar, on its own, could be described as mere courtesy or mere operations. Together, and aimed at a purpose, they become an instrument. The host uses protocol to signal respect, communication to build trust, and logistics to demonstrate competence, and the combined effect shapes how the guest — and the watching world — judges the host nation. This is the practical meaning of #soft_power and reputational security in the setting of a visit: the host is not only being polite but is actively managing perception and relationship (Nye, 2021; Cull, 2022). The literature reviewed here supports this integrated reading from several directions. Work on public and cultural diplomacy shows that reputation is a strategic asset built partly through such encounters (Cull, 2022; Cull, 2023; Zhu and Wyszomirski, 2023). Work on cultural intelligence shows that the ability to adapt across cultures is a learnable skill that improves outcomes in high-stakes cross-cultural settings (Bo and Onwubuya, 2022). Work on gastrodiplomacy shows how a single cultural resource can be turned into an instrument of influence, while also warning that intent and effect are not the same and that the audience's experience is what finally counts (Li and Mok, 2025; Muljabar, 2024). Work on luxury hospitality and hospitableness shows what a premium guest experience is made of and how it can be designed (Christodoulides and colleagues, 2025). And work on event management and sustainability supplies the operational and ethical grounding for the logistics pillar (Santos and colleagues, 2023; Mair and Smith, 2021). Read together, these strands point to the same conclusion: hosting is strategic, it is cross-disciplinary, and it must be integrated to succeed. Implications for practice and education Several practical implications follow for those who plan official visits and for those who train them. First, treat hospitality as strategy, not as background service. The team that plans a visit should begin by asking what the visit is meant to achieve and what image the host wishes to project, and then design protocol, communication, and logistics to serve those goals. This turns a checklist of tasks into a coherent message. Second, integrate the three pillars from the start. A single lead office should hold the whole plan, with protocol, communication, security, and logistics represented around one table and one shared schedule. The seams between teams are where visits fail, so managing those seams is a first-order task, not an afterthought. Third, invest in cross-cultural communication as deliberately as in logistics. Study the guests' communication norms before they arrive, brief principals on the nonverbal expectations they will meet, prepare interpreters thoroughly with glossaries and advance briefings, and build in ways to check that messages are landing as intended. Cultural knowledge should be used as a tested first guess, never as a rigid stereotype (Bo and Onwubuya, 2022). Fourth, design the guest experience around hospitableness. Borrow from the standards of #luxury_hospitality — anticipation of needs, consistency across every contact, and genuine warmth — while remembering that in this setting comfort must be balanced with security. Train every member of the team, down to the driver and the greeter, as a representative of the nation. Fifth, plan for the unexpected. Identify the fragile points of a visit, build alternatives for each, and rehearse who decides when something goes wrong. The aim is not a perfect day but a day whose problems the guest never sees. Sixth, take sustainability seriously, both because it reduces real harm and because it now carries a diplomatic message (Santos and colleagues, 2023; Mair and Smith, 2021). Practical steps at the banquet and in transport are visible signs of shared values. For education, the implication is a call for programs that cross the old boundaries. A course of study in #diplomatic_hospitality would combine the formal knowledge of diplomacy and protocol, the theory and skill of cross-cultural communication and cultural intelligence, and the operational discipline of event and hospitality management, taught through simulations and real case studies. Such an education would prepare graduates not only for foreign services but for the growing number of organizations — international bodies, event companies, tourism boards, and large corporations — that host high-level guests and need the same blended skill set. Limitations and directions for future research This article has limits that shape how its conclusions should be read. As a narrative review, it interprets and combines existing work rather than testing claims with new data, so it can show that practitioners and scholars believe hospitality shapes outcomes but cannot measure how much. The available research is also uneven. Studies of #gastrodiplomacy, for example, cluster around a small number of mostly Asian cases and tend to study what senders intend rather than what audiences feel (Li and Mok, 2025; Muljabar, 2024), and the same imbalance appears in other strands of the literature, which lean toward a handful of well-documented countries and traditions. Much of the deepest practical knowledge in this field lives in government manuals and in the experience of serving officials, and it rarely reaches academic journals, which means scholarship sees only part of the picture. These gaps point to a rich agenda for future research. First, the field would benefit from studies that follow specific visits closely and connect hospitality decisions to measurable outcomes, so that the link between welcome and result rests on evidence rather than belief. Second, more attention should go to the guest's side: how delegations actually experience the hospitality offered to them, and whether the host's intended message is the one received. Third, the field needs broader geographic and cultural coverage, so that the customs of many traditions are studied rather than a favored few. Fourth, the fast changes brought by digital and hybrid formats, by rising security threats, and by sustainability expectations all deserve focused study, since practice is moving faster than research. Finally, because #diplomatic_hospitality is inherently cross-disciplinary, the most useful future work will be collaborative, drawing together scholars of diplomacy, communication, security, and hospitality to study a subject that none of them owns alone. Conclusion The reception of a foreign guest of state looks, from the outside, like a matter of ceremony and comfort. This article has argued that it is much more: it is a working instrument of statecraft, in which #protocol, #cross_cultural_communication, and #logistics combine to shape how nations perceive and deal with one another. The red carpet, the seating chart, the menu, the motorcade, and the interpreter's careful word are not decoration around the real business of diplomacy; on many occasions, they are part of the business itself. Three ideas have run through the discussion. #Diplomatic_hospitality is a single field with three connected pillars that must be integrated rather than handled in isolation. It is a form of #soft_power and #reputational_security, because what a host does with welcome, respect, and care sends signals that a global audience reads. And it is a skilled profession that deserves proper training, one that blends the knowledge of the diplomat, the sensitivity of the communicator, and the discipline of the event manager. For students, the practical message is that the small details of hosting carry large meaning, and that learning to manage them well is learning a genuine craft of international relations. For practitioners, the message is that hospitality is worth planning as carefully as policy, because a warm, competent, and culturally intelligent welcome can build the trust on which everything else depends. In a crowded and closely watched world, the way a nation says "welcome" is one of the clearest statements it makes about who it is. References Bo, M., and Onwubuya, G. C. (2022). Commentary: Intercultural competence past, present, and future: Respecting the past, problems in the present, and forging the future. Frontiers in Psychology, 13, 836372. https://doi.org/10.3389/fpsyg.2022.836372 Christodoulides, G., Chatzipanagiotou, K., Baker, J., and Buhalis, D. (2025). Conceptualizing and measuring customer luxury experience in hotels. Journal of Travel Research. https://doi.org/10.1177/00472875251363846 Cull, N. J. (2022). From soft power to reputational security: Rethinking public diplomacy and cultural diplomacy for a dangerous age. Place Branding and Public Diplomacy, 18(1), 18-21. https://doi.org/10.1057/s41254-021-00236-0 Cull, N. J. (2023). The war for Ukraine: Reputational security and media disruption. Place Branding and Public Diplomacy, 19(2), 195-199. https://doi.org/10.1057/s41254-022-00281-3 Dubinsky, Y. (2023). Clashes of cultures at the FIFA World Cup: Reflections on soft power, nation building, and sportswashing in Qatar 2022. Place Branding and Public Diplomacy. https://doi.org/10.1057/s41254-023-00311-8 Li, G., and Mok, K. (2025). More than two decades of gastrodiplomacy: A review of the concept, current characteristics, and future trends. Frontiers in Political Science, 7, 1659162. https://doi.org/10.3389/fpos.2025.1659162 Mair, J., and Smith, A. (2021). Events and sustainability: Why making events more sustainable is not enough. Journal of Sustainable Tourism, 29(11-12), 1739-1755. Muljabar, H. (2024). Research of trend gastrodiplomacy in Indonesia. Journal of Islamic World and Politics, 8(1), 85-93. https://doi.org/10.18196/jiwp.v8i1.72 Nye, J. S. (2021). Soft power: The evolution of a concept. Journal of Political Power, 14(1), 196-208. https://doi.org/10.1080/2158379X.2021.1879572 Octastefani, T., and Kusuma, B. M. A. (2022). Exploring Taiwanese street food in contemporary Indonesian society: Between nostalgia and gastrodiplomacy. Journal of Government, 7(4), 797-809. https://doi.org/10.31506/jog.v7i4.15663 Putri, A. R., and Baskoro, R. M. (2023). A "splash" of Taiwan bubble tea: Gastrodiplomacy, nation brand and a symbol. AEGIS Journal of International Relations, 7(1), 75-88. https://doi.org/10.33021/aegis.v7i1.4624 Santos, J. A. C., Fernandez-Gamez, M. A., Guevara-Plaza, A., Custodio Santos, M., and Pestana, M. H. (2023). The sustainable transformation of business events: Sociodemographic variables as determinants of attitudes towards sustainable academic conferences. International Journal of Event and Festival Management, 14(1), 1-22. https://doi.org/10.1108/IJEFM-05-2022-0041 Yayusman, M. S., and Mulyasari, P. N. (2024). Indonesia's spice-based gastrodiplomacy: Australia and Africa continents as the potential markets. Journal of ASEAN Studies, 12(1), 51-77. https://doi.org/10.21512/jas.v12i1.8004 Yayusman, M. S., Yaumidin, U. K., and Mulyasari, P. N. (2023). On considering Australia: Exploring Indonesian restaurants in promoting ethnic foods as an instrument of Indonesian gastrodiplomacy. Journal of Ethnic Foods, 10, 1-14. https://doi.org/10.1186/s42779-023-00207-1 Zhu, B., and Wyszomirski, M. J. (2023). Designing cultural diplomacy policy: Structuring a flagship mechanism. International Journal of Cultural Policy, 29(5), 558-575. https://doi.org/10.1080/10286632.2022.2092102 Hashtags #Diplomatic_Hospitality #Protocol #Cross_Cultural_Communication #Logistics_Management #Soft_Power #Public_Diplomacy #Cultural_Diplomacy #High_Level_Delegations #State_Visit #Diplomatic_Protocol #Precedence #Cultural_Intelligence #Gastrodiplomacy #Event_Management #Reputational_Security

For more than a century, the company has been treated in law as a person separate from the people who own and run it. That principle, known as separate legal personality, made modern business possible, but it also gave criminals a way to hide. When ownership can be layered across several countries and buried under nominee directors, the individual who really controls a company and really benefits from its money can disappear from view. This article examines how governments are now trying to reverse that concealment through registers of #ultimate_beneficial_ownership, and it argues that these registers have become the leading legal instrument in the fight against #shell_companies and #trade_based_money_laundering. The study uses a doctrinal and comparative method, drawing together the revised standard set by the Financial Action Task Force, the United States Corporate Transparency Act and its sharp reversal in 2025, the European Union framework after the Court of Justice struck down public access in 2022, and the United Kingdom reforms centred on Companies House. The article explains what #beneficial_ownership is, how the #corporate_veil is used to obstruct investigators, and how registries function as a regulatory way of looking behind that veil. It then evaluates the tensions that keep these tools from working as intended, including the clash between #transparency and #privacy, weak data verification, uneven enforcement, and the risk that criminals simply move to friendlier jurisdictions. The central finding is that registries are necessary but not sufficient: their value depends almost entirely on whether the data inside them is accurate, verified, and actually used. Keywords: beneficial ownership; corporate veil; shell companies; trade-based money laundering; anti-money laundering; beneficial ownership registries; financial crime; corporate transparency 1. Introduction Money that comes from crime rarely announces itself. Drug proceeds, bribes, stolen state assets, and the profits of fraud all need to be moved, cleaned, and made to look ordinary before they can be spent or invested. The people who move this money have learned that the safest hiding place is not a suitcase of cash or a numbered account, but a company. A company can open bank accounts, sign contracts, own property, and send invoices, all while keeping the identity of the human being behind it out of sight. This is why the study of #financial_crime keeps returning to a single question: who really owns and controls the company in front of us. The scale of the problem is hard to measure precisely, which is part of the problem itself. The United Nations Office on Drugs and Crime has long estimated that the amount of money laundered globally each year sits somewhere between two and five percent of world economic output, a range that translates into hundreds of billions and possibly trillions of United States dollars. A large share of that flow passes through corporate structures at some point. When the International Consortium of Investigative Journalists released the Panama Papers, the leak exposed roughly two hundred and fourteen thousand offshore entities and became one of the clearest public demonstrations of how ownership can be hidden at scale (Kejriwal, 2021). Later leaks, including the FinCEN Files and the Pandora Papers, showed that the pattern was not a one-off but a system (Aristodemou, 2024). Two abuses sit at the centre of this article. The first is the #shell_company, a legal entity that exists mainly on paper, with little or no real activity, whose purpose is to hold value and obscure its origin. The second is #trade_based_money_laundering, often shortened to TBML, in which the movement and pricing of goods across borders is manipulated so that value can be shifted between countries under the cover of ordinary commerce. These two abuses reinforce each other. A shell company gives a launderer a face to present to a bank or a customs office, and a trade transaction gives that shell company a reason to be sending or receiving large sums. Behind both stands the same obstacle for investigators, which is the #corporate_veil that separates the company from the individual who profits from it. The response that most governments have converged on is deceptively simple. If criminals rely on hiding the natural person behind a company, then the law should require that person to be identified, recorded, and, where appropriate, disclosed. A register of beneficial owners is the practical form of that idea. The Financial Action Task Force, the global standard setter for #anti_money_laundering policy, rewrote its rules in 2022 to push every country toward this model (FATF, 2022). The United States, the European Union, and the United Kingdom have all built or rebuilt registers in the years since. Yet the path has been anything but smooth. A court in Europe found that giving the general public open access to these registers went too far and breached fundamental rights (WM and Sovim SA v Luxembourg Business Registers, 2022). In the United States, a landmark law was gutted by administrative action within a year of taking effect (FinCEN, 2025). These reversals make the topic timely and contested rather than settled. This article has four aims. First, it clarifies the core concepts, because #beneficial_ownership, the #corporate_veil, shell companies, and TBML are often described loosely. Second, it sets out the global standard and then compares how three major jurisdictions have translated it into law. Third, it explains in concrete terms how a #beneficial_ownership_registry is meant to disrupt shell companies and trade-based schemes. Fourth, it evaluates why registries so often fall short, and what would have to change for them to work. The argument throughout is that a register is only as good as the truthfulness of what is written in it and the willingness of authorities to act on it. 2. Conceptual and Legal Foundations 2.1 What beneficial ownership means The phrase #beneficial_ownership distinguishes between two kinds of owner. The legal owner is the name on the share register or the company documents. The beneficial owner is the natural person who ultimately owns or controls the entity and who ultimately enjoys its benefits, even where that control is exercised through other companies, trusts, nominees, or informal arrangements. The standard-setting definition adopted internationally focuses on the real human being at the end of the chain, not on the intermediate legal shells (FATF, 2023). Most systems now use an ownership threshold to identify who counts. A common test is that any individual who holds or controls a set percentage of shares or voting rights is treated as a beneficial owner. The European Union has settled on a threshold of twenty-five percent or more of ownership interest, held directly or indirectly, with power for the European Commission to lower that figure to fifteen percent in higher-risk sectors (Regulation (EU) 2024/1624). Thresholds are useful because they are administrable, but they are also a weakness, because ownership can be deliberately split among associates so that no single person crosses the line. For this reason the definition also captures control exercised by other means, such as the right to appoint or remove directors, and it provides a fallback: where no natural person can be identified through ownership, the senior managing official may be recorded instead. The idea of #beneficial_ownership also appears in tax law, where it governs who is entitled to treaty benefits on income such as dividends and interest, and where the meaning has generated its own body of dispute (Kuzniacki, 2025). The anti-money-laundering meaning and the tax meaning are related but not identical, and conflating them causes confusion. In this article the focus is the anti-money-laundering sense: identifying the person a criminal would most want to keep hidden. 2.2 The corporate veil and separate legal personality The #corporate_veil is a metaphor for a rule of company law. Since the separation of the company from its members became firmly established in common law, a company has been treated as a distinct legal person, so that its debts are its own and its members are shielded from personal liability. This doctrine is the engine of enterprise. It allows people to invest and to take business risks without betting their homes, and it lets companies act in the market as stable, continuous actors. The same shield, however, can be abused. Courts have always recognised narrow situations in which they will look behind the company and treat its acts as the acts of the people controlling it. This is called piercing or lifting the veil. The doctrine remains tightly limited and its exact boundaries are contested across jurisdictions, with scholars distinguishing between concealment and genuine evasion of legal duties (Lo, 2023; Lauar, 2023). Traditional veil-piercing is a judicial remedy, applied after the fact, in a particular dispute, usually to hold an individual liable or to stop an abuse of the corporate form. The insight that drives modern transparency policy is that traditional veil-piercing is too slow and too narrow to deal with #money_laundering. A launderer does not wait for litigation. By the time a court might pierce the veil, the money has moved on and the shell has been dissolved. What investigators need is not an occasional judicial remedy but continuous, upfront visibility of who stands behind every company. A #beneficial_ownership_registry can be understood as a regulatory form of veil-piercing that operates in advance and at scale. Instead of a judge lifting the veil in a single case, the law requires the veil to be at least partly transparent from the moment a company is formed. 2.3 Shell companies as instruments of concealment A #shell_company is not illegal in itself. Businesses use dormant or holding entities for many legitimate reasons, including preparing for a future venture, holding intellectual property, or structuring a group. The problem is that the very features that make shells convenient for legitimate use, meaning low cost, minimal activity, and thin disclosure, also make them ideal for hiding illicit value. Recent work using data analytics to separate legitimate from illegitimate shells shows how difficult the distinction can be, because on paper a criminal shell and a lawful holding company can look almost the same (Tiwari, Gepp and Kumar, 2025). Launderers exploit shells across the recognised stages of #money_laundering. In the placement stage, illicit funds enter the financial system, often through a company account that gives them a veneer of commercial origin. In the layering stage, funds are moved between entities and across borders to break the audit trail, and here chains of shells in multiple countries are especially useful. In the integration stage, the now-disguised funds return to the owner as apparently legitimate business income, loans, or investment returns. At each stage the shell serves the same function: it puts a corporate name where a human name would otherwise be, and it multiplies the layers an investigator must peel back. The most dangerous shells are those formed in places that ask few questions and record little, then used in places with more developed financial systems. Analysis of leaked datasets has repeatedly shown that intermediaries such as company formation agents and certain professional advisers play a central role in building and operating these structures (Kejriwal, 2021). This is why the transparency agenda targets not only companies but also the professionals who create them, and why several reforms now impose #due_diligence duties on the corporate service providers themselves. 2.4 Trade-based money laundering Of all the ways to launder money, #trade_based_money_laundering is among the hardest to detect, because it hides inside the vast, legitimate flow of world trade. The core technique is to misrepresent the price, quantity, or quality of goods so that value moves from one party to another without an obvious transfer of money. Systematic review of the field identifies several recurring methods: over-invoicing and under-invoicing, where the stated price of goods is set above or below their true value; multiple invoicing, where the same shipment is billed more than once; over-shipment and under-shipment, where the quantity shipped does not match the quantity invoiced; and false description of goods, where cheap items are dressed up as expensive ones or vice versa (Tiwari, Ferrill and Allan, 2025). The estimated magnitude is enormous. One figure cited in the scholarly literature places annual TBML within the European Union alone at between roughly nine hundred billion and one point eight trillion United States dollars, a range that underlines both the scale of the threat and the uncertainty around measuring it (Tiwari, Ferrill and Allan, 2025). The schemes have grown more sophisticated over time, adapting to each new control in a pattern that scholars have described as a regulatory dialectic, in which regulators and launderers respond to one another in successive moves (Sinno et al., 2023). The link between TBML and #beneficial_ownership is direct. A trade-based scheme needs counterparties: an exporter to raise the invoice and an importer to pay it, or a web of both across several countries. Shell companies fill those roles. If the true owner of the exporter and the true owner of the importer are the same person or the same network, then what looks like an arm's length trade is in reality a self-dealing transfer of value. An investigator who can see the beneficial owners on both sides of the trade can spot the connection. An investigator who sees only two unrelated company names cannot. This is the precise point at which a #beneficial_ownership_registry becomes a weapon against TBML: it lets the authorities connect the parties to a trade to the people behind them. 3. Method and Scope This article uses a doctrinal and comparative legal method, supported by a review of recent peer-reviewed literature. The doctrinal element examines primary sources, meaning the international standard, statutes, regulations, and a leading court judgment, to establish what the law requires. The comparative element sets three jurisdictions side by side, the United States, the European Union, and the United Kingdom, chosen because each is a major financial centre, each has recently reformed its rules, and together they display the full range of design choices, from public registers to authority-only access to administrative reversal. The literature review draws on scholarship published mainly within the last five years to situate the legal analysis in current academic debate. The scope is deliberately limited in three ways. First, the article concentrates on legal persons, primarily companies, rather than the parallel and equally important rules on trusts and other legal arrangements, although these are noted where relevant. Second, it treats registries as the central case study rather than attempting a full survey of every #anti_money_laundering tool. Third, because the law in this area is changing quickly, the analysis reflects the position as understood at the time of writing and should be read with that caveat in mind. The aim is not to predict the outcome of ongoing reforms but to explain the logic of the tools and to assess their strengths and weaknesses. 4. The Global Standard: FATF Recommendation 24 Any serious discussion of #beneficial_ownership begins with the Financial Action Task Force, usually called the #FATF. It is an inter-governmental body that sets the recommendations most countries follow on combating money laundering and the financing of terrorism. Its recommendations are not treaties, but they carry great weight, because countries are assessed against them and a poor assessment can damage a country's standing in the international financial system. The relevant standard is Recommendation 24, which deals with transparency and #beneficial_ownership of legal persons. For years this recommendation allowed countries considerable latitude, and successive rounds of country assessments showed that many jurisdictions struggled to give their authorities reliable, up-to-date ownership information. In March 2022 the FATF adopted a significant revision (FATF, 2022). The revised standard moved the world toward a stronger and more consistent model, and its main features are worth stating plainly because national laws now echo them. First, the revised Recommendation 24 requires countries to ensure that competent authorities have access to information that is adequate, accurate, and up to date on the true owners of companies, and that this access is timely. The three adjectives matter. Adequate means enough to identify the natural person and the way they exercise control. Accurate means verified rather than simply self-declared. Up to date means maintained over time, not frozen at the moment of incorporation. Second, and most importantly for this article, the standard requires that #beneficial_ownership information be held by a public authority or body functioning as a #beneficial_ownership_registry, or by an alternative mechanism that provides equally efficient access. This was the decisive shift. Where the older approach let companies simply hold their own information, the revised approach pushes countries toward a central register as the default. The FATF also endorsed a multi-pronged approach, combining a register with information held by financial institutions and other authorities, on the evidence that countries using several sources at once were more effective than those relying on a single source (FATF, 2023). Third, the revision tightened the rules on tools that had long been used to obscure ownership. It called for prohibiting the issuance of new bearer shares, which are shares owned by whoever physically holds the certificate and therefore almost impossible to trace, and for immobilising or converting existing ones. It also required stronger transparency around nominee arrangements, in which one person holds shares or acts as director on behalf of a hidden principal (FATF, 2022). In February 2023 the FATF made parallel changes to Recommendation 25, which covers trusts and similar legal arrangements, so that the standards for arrangements would broadly match those for companies (FATF, 2023). Together the two revisions represent the clearest global consensus yet that concealment of #beneficial_ownership is a first-order threat and that registries are the primary answer. The FATF has reported that its work in this area led numerous countries to commit to establishing registries, and its next round of country assessments places greater emphasis on whether measures actually work rather than merely whether they exist on paper. It is worth stressing what the FATF standard does not do. It does not require that registers be open to the general public. It requires timely access for competent authorities and, in practice, for regulated businesses conducting #due_diligence. Whether to go further and open the register to journalists, civil society, or anyone at all is left to each jurisdiction. That single design choice, public access or not, has become the most contested question in the entire field, and it is the fault line along which the European experience fractured. 5. Comparative Analysis of National and Regional Regimes 5.1 United States: the Corporate Transparency Act and its reversal For a long time the United States was, in the language of transparency campaigners, one of the easiest places on earth to form an anonymous company. States such as Delaware and Wyoming allowed companies to be created quickly and cheaply without collecting information on the human beings behind them. The result was that United States shells appeared throughout international laundering schemes, an irony given the country's aggressive enforcement of its own financial laws abroad. The #Corporate_Transparency_Act, enacted at the start of 2021 as part of a broader defence spending law, was meant to close this gap. It required a large class of companies, described as reporting companies, to file information about their beneficial owners with the Financial Crimes Enforcement Network, known as #FinCEN. The information was not to be a public register in the European sense but a confidential database available to law enforcement and, under safeguards, to financial institutions. Reporting obligations took effect on the first of January 2024, and for the first time millions of small United States entities were required to name the individuals who owned or controlled them (Aristodemou, 2024). What followed was one of the most striking reversals in the recent history of #anti_money_laundering policy. The law faced sustained legal challenge and a series of shifting deadlines through late 2024 and early 2025. Then, in the spring of 2025, the executive branch changed course. In February 2025 #FinCEN announced it would not enforce penalties against companies that failed to file by the looming deadline. On the second of March 2025 the Treasury Department signalled a narrowing of the rule. On the twenty-first of March 2025 FinCEN issued an interim final rule, formally published on the twenty-sixth of March, that redefined the term reporting company to mean only entities formed under the law of a foreign country and registered to do business in the United States (FinCEN, 2025). In one stroke, all entities created within the United States, together with their beneficial owners, were exempted from the obligation to report. United States persons were also relieved from having to provide their information even where they are beneficial owners of a foreign reporting company. The practical effect is that the domestic core of the law, the part aimed at anonymous shells formed inside the United States, was switched off. Only foreign companies registered to do business in the country remain within scope, and even they need not report their United States-person owners. The government justified the change as a reduction of regulatory burden on small businesses. Critics pointed out that it removed precisely the transparency the statute was designed to create. The legal picture became more complicated still. In December 2025 a federal appellate court held that the #Corporate_Transparency_Act is a constitutional exercise of legislative power and does not, on its face, violate constitutional protections. Yet that ruling addressed the validity of the statute, not the administrative rule that had narrowed it. The consequence is a curious split: the law stands as constitutional, while the regulation implementing it exempts most domestic companies from actually complying. For the study of registries, the United States episode is a cautionary tale. It shows that even a landmark statute, hard-won over years, can be hollowed out by administrative action, and that the durability of a transparency regime depends on sustained political commitment as much as on the words in the statute book. 5.2 European Union: from public registers to the Sovim ruling and the 2024 package The European Union moved earlier and, for a time, further than the United States. Its Fourth Anti-Money Laundering Directive of 2015 required member states to set up central registers of #beneficial_ownership and to allow access to persons who could show a legitimate interest. The Fifth Anti-Money Laundering Directive of 2018 went further and required that these registers be open to any member of the general public. For campaigners this was a high-water mark. Public registers meant that journalists, researchers, and ordinary citizens could look up who owned a company, and cross-border investigations that once took months could be advanced in minutes. That model collided with European fundamental rights law. In two joined cases arising from Luxembourg, a company and a beneficial owner asked to have their information restricted, arguing that open public access to their data breached their rights to private life and to protection of personal data. On the twenty-second of November 2022 the Court of Justice of the European Union, sitting as its Grand Chamber, agreed. It held that the provision of the directive requiring that #beneficial_ownership information be accessible to any member of the general public was invalid, because the interference with the rights guaranteed by the Charter of Fundamental Rights was neither limited to what was strictly necessary nor proportionate to the aim pursued (WM and Sovim SA v Luxembourg Business Registers, 2022). The reaction was immediate and disruptive. Several member states, including Luxembourg, the Netherlands, and Ireland, suspended public access to their registers within days. For a period, investigators, journalists, and even businesses conducting #due_diligence found that a tool they had come to rely on was switched off or sharply curtailed (Thomas-James, 2023). The judgment was widely read as a serious setback for #corporate_transparency, and some commentators framed it as the possible beginning of the end for the open-register model in Europe (Thomas-James, 2023). Others noted that the Court had not condemned registers as such, and had accepted that #legitimate_interest access, including by the press and civil society, remained lawful; the problem was unrestricted access by anyone at all. Into this uncertainty the European Union delivered a comprehensive reform. In 2024 it adopted a package of measures that reshape the entire framework. The package rests on three main instruments. The first is the Anti-Money Laundering Regulation, Regulation (EU) 2024/1624, a single rulebook that applies directly in every member state without the need for national transposition, harmonising #due_diligence and #beneficial_ownership rules that had previously varied from country to country (Regulation (EU) 2024/1624). The second is the Sixth Anti-Money Laundering Directive, Directive (EU) 2024/1640, which governs the institutional side, including the design of national registers, the powers of financial intelligence units, and the rules on access (Directive (EU) 2024/1640). The third is the regulation establishing a new Authority for Anti-Money Laundering and Countering the Financing of Terrorism, the body known as #AMLA, seated in Frankfurt and operational since the first of July 2025, which will directly supervise a group of the highest-risk financial institutions from 2028. Several features of the package respond directly to the weaknesses the earlier system had revealed. On the definition of ownership, the threshold is set at twenty-five percent or more, closing a small gap in the previous wording and allowing the Commission to lower it to fifteen percent for higher-risk categories (Regulation (EU) 2024/1624). On access, the new directive re-establishes access for persons with a #legitimate_interest, expressly including journalists and civil society organisations, with more harmonised criteria across the Union, an approach designed to satisfy the Court while preserving meaningful scrutiny. On interconnection, national registers are to be linked across the Union so that a beneficial owner can be traced across borders through a single system rather than checked country by country. The bulk of these obligations apply from the tenth of July 2027, with some register-access provisions phased in earlier. The reform also standardises the data each register must hold, so that names, dates of birth, nationalities, countries of residence, and the nature and extent of each interest are recorded in a common format, which is intended to make the information easier to reconcile and to use (Matras, 2025). The European story therefore contains both a warning and a lesson. The warning is that a transparency measure designed without adequate attention to privacy and proportionality can be struck down, taking a working tool offline in the process. The lesson is that the response need not be retreat: the 2024 package attempts to rebuild the system on a firmer footing that balances #transparency against #data_protection rather than sacrificing one for the other. Whether member states will implement it well, and whether the harmonised #legitimate_interest test will prove workable in practice, are open questions that recent scholarship treats with caution (Matras, 2025). 5.3 United Kingdom: Companies House reform and the Register of Overseas Entities The United Kingdom occupies an interesting middle position. It was an early adopter of a public register of company ownership, through its People with Significant Control regime introduced in 2016, which made information about controllers of United Kingdom companies openly available. Because this is a matter for domestic law rather than European Union law, the United Kingdom register was not directly affected by the Court of Justice ruling, and it has remained public. Two more recent statutes deepened the framework. The first, the Economic Crime (Transparency and Enforcement) Act 2022, created a #Register_of_Overseas_Entities, which requires foreign entities that own certain interests in United Kingdom land to disclose their beneficial owners. The measure was rushed through in response to concerns about hidden foreign wealth in the property market. Its early record was mixed. One study found that by early March 2023 only about twenty-six thousand of roughly one hundred and thirty-nine thousand entities that were required to register had actually done so, a compliance rate of under nineteen percent, which illustrates how a register can exist on paper long before it captures the population it is meant to cover (Aristodemou, 2024). The second and more far-reaching statute is the Economic Crime and Corporate Transparency Act 2023, which received Royal Assent in October 2023 and is being brought into force in stages (Economic Crime and Corporate Transparency Act 2023). Its central idea is to transform Companies House, the United Kingdom company registrar, from a passive filing cabinet into an active gatekeeper. Historically the registrar accepted documents largely on trust and placed them on the public record with little checking, which allowed fraudsters to register fictitious directors and to use false addresses. Under the new law the registrar gains powers to query, reject, and remove information, and to share intelligence with law enforcement. The most consequential change for the concealment problem is mandatory #identity_verification. From the eighteenth of November 2025 the regime requires directors and people with significant control to verify their identity, with new appointees verifying before they can act and existing officeholders verifying through a transition period, using either a government digital service or an authorised corporate service provider that is itself supervised for #anti_money_laundering purposes. The point is straightforward: a register of owners is worth little if the names in it are invented, so the law attacks the problem at the source by insisting that the human beings behind a company be real and identified. Early operational reporting from Companies House indicates the scale of the clean-up already under way, with tens of thousands of false registered office addresses, officer addresses, and control records removed since the first powers took effect, and thousands of fraudulent appointments stripped from the register in targeted investigations. These figures suggest both how much abuse the old system tolerated and how a more active registrar can begin to correct it. Taken together, the three jurisdictions display the full spectrum of choices. The United Kingdom keeps a public register but concentrates on making its data true through verification. The European Union rebuilds toward interconnected registers with #legitimate_interest access after a constitutional setback. The United States built a confidential database and then, through administrative action, exempted most of the companies it was meant to cover. The common thread is that every system is grappling with the same trade-offs of access, accuracy, and enforcement, and none has solved them completely. 6. How UBO Registries Work Against Shell Companies and TBML Having set out the standard and the national regimes, it helps to explain in practical terms how a #beneficial_ownership_registry actually disrupts the abuses at the centre of this article. The mechanism operates through several linked effects. The first effect is identification. At the moment a company is formed, and on an ongoing basis, the law requires that the natural person or persons who own or control it be recorded. This alone changes the calculus for a would-be launderer. Where forming an anonymous shell once cost little and revealed nothing, it now requires putting a real name against the entity. For criminals this raises the risk of exposure, and for the honest majority it imposes only a modest administrative step. The register turns the default from anonymity to attribution. The second effect is connection. A single company record is useful, but the real power of a register appears when records can be searched and linked. If an investigator can query a register by the name of an individual, they can find every company that person controls. This is what allows the connection of parties in a #trade_based_money_laundering scheme. Recall that TBML typically relies on an exporter and an importer, often across borders, whose invoices move value while appearing to be ordinary trade. If the beneficial owner of the exporter and the beneficial owner of the importer turn out to be the same person or the same closely linked network, the apparent trade is revealed as a self-dealing transfer. Without the register, the two companies are just two names; with it, they can be tied to a common controller. This is precisely why the European Union has emphasised interconnection of national registers, since TBML is inherently cross-border and a register that stops at the national frontier can be defeated simply by placing one side of the trade abroad. A simple illustration makes the connection effect concrete. Imagine a consignment of electronics that is invoiced by a company in one country at ten times its real market value and paid in full by a company in another country. On the surface this is a normal, if expensive, trade. The importer has paid, the exporter has been paid, and the goods have crossed a border. Nothing in the shipping documents looks wrong. Now suppose an investigator queries a register and discovers that both the exporter and the importer are ultimately controlled by the same individual, or by two relatives acting together. The picture changes at once. What appeared to be an ordinary sale is revealed as a device for moving nine parts value out of one country and into another under cover of a one part real transaction. The over-invoicing is the laundering method, and the shared beneficial owner is the fingerprint. Without the register, the investigator sees two unrelated firms and a plausible invoice. With it, the investigator sees a single controller sitting on both sides of the deal. This is the everyday work that ownership transparency is meant to enable, and it explains why the value of a register grows sharply once it can be searched by person and linked across borders rather than merely read one company at a time. The third effect is due diligence. Registers do not only serve the state. Banks, lawyers, accountants, and other regulated businesses are obliged to know their customers, and a central register gives them a source against which to check what a customer tells them. If a client claims a certain ownership structure and the register shows something different, that discrepancy is a red flag that can trigger further inquiry or a report to the authorities. Empirical work on customer due diligence confirms that the quality of ownership information is central to whether these checks catch anything (Gaviyau and Sibindi, 2023). In the European framework, the obligation on regulated businesses to consult the register and to report mismatches is meant to create a feedback loop that improves the register's own accuracy over time. The fourth effect is deterrence and disruption. A shell company is valuable to a criminal only so long as it hides the owner. Once ownership is recorded and potentially discoverable, the shell becomes a liability rather than an asset, because it now carries a documentary link back to the person it was meant to protect. Where identity verification is layered on top, as in the United Kingdom reform, the criminal cannot even populate the register with a convincing false name. The combined result is to raise the cost and lower the reliability of the concealment techniques on which shells and TBML depend. It is important to be honest about the limits of these effects, which the next section examines in detail. A register identifies owners only if the reported information is truthful; it connects parties only if it is searchable and linked; it supports due diligence only if regulated businesses actually use it; and it deters only if there is a credible prospect of enforcement. Each of these conditions can fail. The mechanism is sound in theory, but its performance in practice is contingent on design and implementation choices that many jurisdictions have not yet got right (Matras, 2025; Tiwari, Gepp and Kumar, 2025). 7. Discussion: Tensions, Weaknesses, and Contested Questions 7.1 Privacy versus transparency The sharpest tension in this field is between #transparency and #privacy. The case for openness is powerful. Public registers let journalists and civil society investigate wrongdoing that overstretched authorities might never reach, and much of what the world knows about grand corruption came from exactly this kind of scrutiny (Aristodemou, 2024). The case for restraint is also serious. A beneficial owner is an identified individual, and publishing their financial connections can expose them to profiling, to unwanted attention, and in some cases to real danger. The European Court weighed these interests and found that blanket public access went too far, because it was not limited to what was strictly necessary (WM and Sovim SA v Luxembourg Business Registers, 2022). The resolution most systems are converging on is a tiered model of access. Competent authorities receive full and timely access. Regulated businesses receive access for #due_diligence. Persons with a demonstrated #legitimate_interest, such as journalists and non-governmental organisations, receive access under defined criteria. The general public may receive limited access or none. The difficulty is defining #legitimate_interest in a way that is neither so narrow that it shuts out genuine watchdogs nor so broad that it recreates the open access the Court rejected. This is not a settled matter, and the workability of the harmonised European test will be one of the important tests of the 2024 reform (Matras, 2025). 7.2 Data quality and verification A register is only as good as the truth of what is inside it. This is the single most important practical weakness of the entire model. If beneficial owners can simply declare whatever they like, with no checking, then a criminal will declare a false name or a compliant nominee, and the register will faithfully record a lie. Scholarship on the European registers finds that a large share of the data entered is questioned by the very businesses and institutions that are meant to rely on it, which erodes the register's usefulness as an #anti_money_laundering tool (Matras, 2025). The nominee problem deserves particular attention, because it is the classic way of defeating a register while still complying with it on paper. A nominee is a person who appears on the record as owner or director but who acts entirely on the instructions of a hidden principal. If the register captures only the nominee, it records a truthful-looking name that points to the wrong person. This is why the revised global standard called for stronger transparency around nominee arrangements, and why several reforms now require nominees to disclose the person on whose behalf they act (FATF, 2022). The same logic extends to the professionals who build corporate structures for a living. Company formation agents, some lawyers, and certain accountants can, wittingly or not, become the architects of concealment, and studies of leaked datasets place these intermediaries at the centre of the most complex structures (Kejriwal, 2021). Modern frameworks respond by treating these service providers as obliged entities in their own right, subject to their own due diligence and reporting duties, so that the people who assemble shells share responsibility for knowing who the real owners are. A register that ignored nominees and enablers would be checking the front door while leaving the side entrance open. This is why verification, rather than mere collection, is the frontier of reform. The United Kingdom approach of mandatory #identity_verification attacks the problem at its root by ensuring that the individuals named are real and that their identities have been checked against reliable evidence. The revised FATF standard likewise insists that information be accurate, meaning based on verification, not simply adequate on its face (FATF, 2022). Emerging techniques, including data analytics that compare a company's declared profile against patterns typical of illegitimate shells, may help authorities focus verification effort where it is most needed (Tiwari, Gepp and Kumar, 2025). Artificial intelligence tools are being explored to sift large volumes of corporate data for anomalies, though these tools raise their own questions of accuracy and oversight (Garcia-Segura, 2024). None of this removes the underlying point: a register without verification is a list of assertions, and assertions can be false. 7.3 Enforcement and the problem of political will Even a well-designed, well-verified register achieves nothing if no one acts on it. Enforcement depends on resources, on inter-agency cooperation, and above all on sustained political commitment. The United States episode is the clearest illustration. A statute that many regarded as historic was, within a year of taking effect, drained of most of its force by an administrative rule, on the stated ground of reducing burden on business (FinCEN, 2025). Whatever one thinks of that policy choice, it demonstrates that transparency regimes are vulnerable to shifts in political priorities, and that the existence of a law is no guarantee of its operation. The United Kingdom experience with the #Register_of_Overseas_Entities points to a quieter enforcement problem, which is compliance. A register whose target population largely ignores it, as the low early registration rate suggested, does not deliver the transparency promised on paper until the authorities compel compliance (Aristodemou, 2024). Enforcement is unglamorous and expensive, and it is often the first thing to be neglected once a headline reform has passed. The lesson across all three jurisdictions is that the register is the beginning of the work, not the end of it. 7.4 Displacement and the limits of registries A final weakness is displacement. #Financial_crime is mobile. If one jurisdiction tightens its rules, criminals can move their structures to another that has not. This is why a purely national register is inherently limited and why the global standard and the European push for interconnection matter so much. As long as some jurisdictions offer easy anonymous incorporation, launderers can route the sensitive part of a structure through those places while presenting a clean face elsewhere. The interconnection of registers, and the extension of coverage to foreign entities with links to a jurisdiction, are attempts to close this gap, but they will only work if adoption is broad. A world of transparent registers with a few opaque holdouts is a world in which the holdouts attract exactly the business the reforms were meant to stop. There is also a deeper limit worth stating. A register identifies who owns a company, but it does not by itself reveal what the company does or where its money comes from. It is a piece of a larger system that also includes transaction monitoring, suspicious activity reporting, customs data analysis, and international cooperation. Against #trade_based_money_laundering in particular, ownership data must be combined with trade and payment data to be effective, because the abuse lies in the mismatch between the declared trade and the real value moved (Tiwari, Ferrill and Allan, 2025; Sinno et al., 2023). Treating the register as a complete solution rather than as one instrument among several is a mistake that invites disappointment. 8. Limitations and Future Directions This article has limitations that shape how its conclusions should be read. Its comparative focus on three advanced jurisdictions gives a clear view of the leading models but says less about developing economies, where registries may face harder constraints of capacity and where the risks from #shell_companies and #trade_based_money_laundering can be especially acute. Research from a range of national settings suggests that the gap between adopting a rule and implementing it well is often widest precisely where resources are thinnest (Gaviyau and Sibindi, 2023; Setyastomo, Mardikaningsih and Sulani, 2024). A fuller comparative study would extend the analysis to these settings. The article is also a snapshot of a moving target. The European package will not be fully in force until 2027, the United States position may change again as a final rule is issued or as litigation continues, and the United Kingdom verification regime is still being rolled out. Any assessment written today is provisional. This is a limitation, but it is also an invitation, because it means the field is unusually open to timely empirical research. Several directions stand out for future work. The first is measurement. There is still no reliable way to say how much illicit value a register actually prevents from being laundered, and better outcome measures would help move the debate beyond the mere existence of registers toward their real effect. The second is verification technology. The most promising practical advance is the shift from self-declaration to verified data, and careful study of which verification methods work, at what cost, and with what error rates would be valuable (Tiwari, Gepp and Kumar, 2025; Garcia-Segura, 2024). The third is the design of #legitimate_interest access, where the European experiment will generate real-world evidence about how to reconcile scrutiny with #data_protection. The fourth is the integration of ownership data with trade and payment data, which is the key to turning registries into an effective tool specifically against TBML rather than a general transparency measure. 9. Conclusion The company that shields its owner is one of the oldest problems in commercial law and one of the newest fronts in the fight against #financial_crime. The #corporate_veil that makes enterprise possible also makes concealment possible, and criminals have exploited that fact through #shell_companies and #trade_based_money_laundering on a scale measured in the hundreds of billions or more. The response that governments have converged on, a register of #ultimate_beneficial_ownership, is best understood as a regulatory way of piercing that veil in advance and at scale, replacing the slow, case-by-case remedy of the courts with continuous, upfront visibility of who really stands behind a company. The evidence gathered here supports a measured conclusion. Registries have become the primary legal instrument against shells and trade-based schemes, and the direction of global policy, set by the revised Recommendation 24 and reflected in the reforms of the European Union and the United Kingdom, is firmly toward them (FATF, 2022; Regulation (EU) 2024/1624; Economic Crime and Corporate Transparency Act 2023). Yet the tool is only as strong as the conditions that make it work. A register identifies owners only if the data is verified, connects parties only if it is searchable and interconnected, supports #due_diligence only if regulated businesses use it, and deters only if enforcement is real and sustained. The European experience shows that a register can be undone by inattention to #privacy, and the United States experience shows that even a landmark law can be hollowed out by a shift in political will (WM and Sovim SA v Luxembourg Business Registers, 2022; FinCEN, 2025). The practical message for students and future practitioners is therefore twofold. First, the register is necessary. Without it, the person behind a shell simply cannot be found in time to matter, and the abuses at the heart of this article go unchecked. Second, the register is not sufficient. Its promise is realised only through verification, access designed to respect fundamental rights, cross-border interconnection, and the unglamorous work of enforcement. Piercing the corporate veil, in the modern regulatory sense, is not an event but a discipline. The registries are built; the harder task, of keeping them true and using them well, is only beginning. #beneficial_ownership #corporate_veil #ultimate_beneficial_ownership #shell_companies #trade_based_money_laundering #UBO_registries #anti_money_laundering #financial_crime #corporate_transparency #FATF_Recommendation_24 #Corporate_Transparency_Act #FinCEN #EU_AML_package #Companies_House #piercing_the_corporate_veil References Aristodemou, M. (2024). Are beneficial ownership laws important? Exploring the impact of Panama, FinCEN, and Pandora Papers on beneficial ownership laws in the UK and the US. Journal of Economic Criminology, 5, 100082. https://doi.org/10.1016/j.jeconc.2024.100082 Court of Justice of the European Union. (2022). WM and Sovim SA v Luxembourg Business Registers, Joined Cases C-37/20 and C-601/20, Judgment of the Grand Chamber, 22 November 2022. ECLI:EU:C:2022:912. Directive (EU) 2024/1640 of the European Parliament and of the Council of 31 May 2024 on the mechanisms to be put in place by Member States for the prevention of the use of the financial system for the purposes of money laundering or terrorist financing (Sixth Anti-Money Laundering Directive). Economic Crime and Corporate Transparency Act 2023, c. 56 (United Kingdom). Financial Action Task Force. (2022). Public statement on revisions to Recommendation 24. FATF, Paris. Financial Action Task Force. (2023). Guidance on beneficial ownership of legal persons. FATF, Paris. Financial Crimes Enforcement Network. (2025). Beneficial ownership information reporting requirement revision and deadline extension for foreign reporting companies (Interim Final Rule). United States Department of the Treasury. Fraiha Granjo, A., Martini, M., & Sipos, G. (2023). Beneficial ownership registers in the EU: Progress so far and the way forward. Publications Office of the European Union, Luxembourg. Garcia-Segura, L. A. (2024). The role of artificial intelligence in preventing corporate crime. Journal of Economic Criminology, 5, 100091. https://doi.org/10.1016/j.jeconc.2024.100091 Gaviyau, W., & Sibindi, A. B. (2023). Anti-money laundering and customer due diligence: empirical evidence from South Africa. Journal of Money Laundering Control, 26(7), 224-238. https://doi.org/10.1108/JMLC-06-2023-0103 Kejriwal, M. (2021). On using centrality to understand importance of entities in the Panama Papers. PLoS ONE, 16(3), e0248573. https://doi.org/10.1371/journal.pone.0248573 Kuzniacki, B. (2025). Determining the beneficial owner of dividend income compared to other items of income in international taxation. Nordic Tax Journal, 2024(s1), 44-71. https://doi.org/10.2478/ntaxj-2024-0004 Lauar, M. (2023). Piercing the corporate veil in Brazilian, Canadian, and Spanish civil law. Cadernos de Dereito Actual, 21, 82-100. Lo, S. H. C. (2023). Nature of corporate veil-piercing and revitalization of the evasion principle. Law Quarterly Review, 139, 436. Matras, T. M. (2025). Harmonisation of the beneficial ownership registers in European Union. Journal of Money Laundering Control, 28(7), 81-94. https://doi.org/10.1108/JMLC-07-2025-0119 Regulation (EU) 2024/1624 of the European Parliament and of the Council of 31 May 2024 on the prevention of the use of the financial system for the purposes of money laundering or terrorist financing (Anti-Money Laundering Regulation). Setyastomo, D., Mardikaningsih, R., & Sulani, S. (2024). Legal implications of beneficial ownership disclosure in closed companies on strengthening the anti-money laundering system in Indonesia. Journal of Social Science Studies, 4(1), 67-74. Sinno, R. M., Baldan, C., & Ferretti, R. (2023). The evolution of trade-based money laundering schemes: a regulatory dialectic perspective. Journal of Financial Crime, 30(5), 1279-1290. Thomas-James, D. (2023). Editorial: The Court of Justice of the European Union and the beginning of the end for corporate transparency? Journal of Financial Crime. Tiwari, M., Ferrill, J., & Allan, D. M. C. (2025). Trade-based money laundering: a systematic literature review. Journal of Accounting Literature, 47(5), 1-26. https://doi.org/10.1108/JAL-11-2022-0111 Tiwari, M., Gepp, A., & Kumar, K. (2025). Using data analytics to distinguish legitimate and illegitimate shell companies. Journal of Economic Criminology, 6, 100123. https://doi.org/10.1016/j.jeconc.2024.100123

The question of whether animals have language is one of the oldest questions in the study of behaviour, and one of the most misunderstood. This article reviews what is currently known about #animal_communication and asks how far the word "language" can reasonably be stretched. It begins by separating communication from language, using the classic design features of human speech as a checklist rather than a verdict. It then examines the main theoretical frameworks used in the field, including #information_theory, #biosemiotics, and the study of #linguistic_laws such as Zipf's law of abbreviation. The core of the article is a set of case studies drawn from recent empirical work: the #honeybee_waggle_dance, the alarm systems of birds and monkeys, the compositional call sequences of Japanese tits, the gestural repertoires of great apes, the coda systems of sperm whales, the song of humpback whales, the name-like calls of elephants and marmosets, and the acoustic emissions of stressed plants. Each case is used to test a specific structural property: reference, #compositionality, #displacement, #vocal_learning, #arbitrariness, and #cultural_transmission. The article then reviews the rapid growth of computational methods, including #passive_acoustic_monitoring and self-supervised #machine_learning models trained on large bioacoustic datasets, and it evaluates the promise and the risk of these tools. A critical section addresses #anthropomorphism, the #Clever_Hans_effect, the failures of the twentieth century #ape_language_projects, and the statistical traps that make it easy to see grammar where none exists. The article closes with the implications of this research for evolutionary theory, for #conservation, and for #animal_welfare and ethics. The conclusion is deliberately cautious: no known animal system has been shown to possess the full open-ended productivity of human language, but several systems possess individual properties once thought to be uniquely human, and the boundary is now drawn by evidence rather than by assumption. Keywords: animal communication; comparative linguistics; bioacoustics; compositionality; vocal learning; animal cognition; machine learning; language evolution 1. Introduction Every human society has stories about talking animals. The idea that a bird, a wolf, or a fish might speak to us is so old that it appears in the earliest recorded literature. What is new is not the interest but the evidence. Over the last two decades, and especially over the last five years, the study of #ethology and #bioacoustics has produced findings that would have seemed like fantasy to earlier generations of scientists. Elephants appear to call each other by something like names. Sperm whales produce click sequences with a combinatorial structure that resembles a small phonetic system. Wild birds combine calls in an order that changes their meaning. Honeybees learn their dances from older bees rather than performing them perfectly from birth. These discoveries have arrived at the same time as a technological shift. Cheap recorders, long-duration tags, drones, and large neural network models have changed what is possible to measure. A single research team can now collect more hours of whale sound in one field season than an entire generation of earlier researchers could analyse in a career. The bottleneck has moved from data collection to interpretation. This creates both an opportunity and a danger. The opportunity is that questions once answered by opinion can now be answered by measurement. The danger is that powerful pattern-finding tools will find patterns whether or not those patterns carry meaning. A model trained to compress dolphin whistles will always produce a compressed representation. Whether that representation corresponds to anything the dolphin intends is a separate question, and it is a question that no amount of computing power can answer on its own. This article is written for students who are approaching the field for the first time, and who need to understand both the excitement and the caution. It has three aims. The first is to define the terms clearly, because most public confusion about #animal_communication comes from sloppy definitions. The second is to survey the empirical evidence across taxonomic groups, using each case to test a specific structural claim. The third is to explain the methods, including the newer computational ones, well enough that a reader can judge a headline for themselves. The central argument is simple. The question "do animals have language?" is a poor question, because it invites a yes or no answer to something that is not a binary property. A better question is: which of the components of human language appear in which animals, in what form, and why? When the question is asked that way, the field becomes tractable, and the answers become genuinely informative about the evolution of our own species. 2. Communication, Signal, and Language: Three Different Things 2.1 Communication is broader than language In the biological sense, communication occurs when a signal produced by one organism changes the behaviour of another organism, in a way that has been shaped by natural selection because of that effect. This definition is deliberately wide. It includes the chemical trail of an ant, the colour change of a cuttlefish, the tail flick of a deer, and the courtship song of a cricket. It also includes cases where the receiver is a different species, and cases where the signal is dishonest. Notice what the definition does not require. It does not require that the sender intends anything. It does not require that the signal carries a concept. It does not require learning. A moth releasing a pheromone is communicating in the biological sense, and there is no reason to think the moth has a plan. 2.2 Language is narrower Human language is a particular kind of communication system with an unusual combination of properties. The most useful summary remains the list of design features proposed in the mid twentieth century, which is still taught because it functions as a checklist. The features that matter most for our purposes are the following. Semanticity. Signals map onto meanings. The sound sequence "water" refers to a substance. #arbitrariness. The form of the signal has no necessary connection to its meaning. Nothing about the sound of the word "water" is wet. #displacement. Speakers can refer to things that are not present in time or space. We can talk about yesterday, about another country, about an imaginary object. Productivity, sometimes called open-ended productivity. Speakers can produce and understand sentences that have never been said before. The set of possible utterances is unbounded. #duality_of_patterning. A small set of meaningless units, such as sounds, is combined into a large set of meaningful units, such as words, which are then combined again into sentences. This double layer is what allows a small inventory to generate an enormous vocabulary. Cultural transmission. The specific system is learned from other members of the community, not fixed genetically. #syntax. There are rules governing how units combine, and those rules change meaning. "The dog bit the man" and "the man bit the dog" use the same words and mean different things. #recursion. Structures can be embedded inside structures of the same type, in principle without limit. No animal system yet described has all of these. This is the honest starting point, and students should hold onto it. But several animal systems have some of them, and it is the pattern of presence and absence that carries the scientific interest. 2.3 Why the checklist is a tool and not a verdict There are two ways to misuse this list. The first is to use it to shut down inquiry, by noting that no animal has all the features and concluding that the topic is closed. The second is to use it to inflate claims, by finding one feature in one species and announcing that the species has language. Both errors share the same mistake. They treat language as a single indivisible thing that a species either has or lacks. Modern comparative work treats it as a bundle of separable capacities, each with its own evolutionary history, each potentially present in different lineages for different reasons. Under this view, finding #compositionality in a small bird does not make the bird a linguist. It tells us that the cognitive machinery for combining signals is available in a brain far simpler than ours, which in turn tells us that this machinery is not the hard part of language evolution. That is a real result, and it is more interesting than a headline. 3. Theoretical Frameworks 3.1 Information theory and the measurement of structure The most widely used quantitative framework treats a communication system as a channel that transmits information. From a recording, researchers can compute #entropy, which measures how unpredictable the next unit is, and mutual information, which measures how much one unit tells you about another unit some distance away. These measures are attractive because they are objective and do not require any assumption about meaning. They also have a serious limitation. Structure is not meaning. A system can have complex statistical structure and still be a simple system, because the structure comes from the physics of the body producing it, from the environment, or from arousal. High entropy does not prove a rich vocabulary. Low entropy does not prove simplicity. Students should treat information theoretic results as descriptions of form, not as evidence of content. 3.2 Linguistic laws in biology A productive recent research programme applies statistical regularities from human language to animal signals. Two are especially important. Zipf's law of abbreviation states that more frequently used items tend to be shorter. In human language, the commonest words are the shortest. The same pattern has now been reported in the vocal repertoires of several species, including primates and cetaceans, and it is generally interpreted as a signature of pressure toward efficiency in the channel. Menzerath's law states that the longer a construction is, the shorter its constituent parts tend to be. Longer words in human speech tend to have shorter syllables. Similar effects have been reported in animal call sequences. A comprehensive review of these regularities across biology argued that these laws are not curiosities but expected consequences of general optimisation pressures acting on any communication system with costs (Semple, Ferrer-i-Cancho, and Gustison, 2022). This is an important framing. The presence of #Zipfs_law in a species does not mean the species has words. It means the species is under the same kind of economic pressure that shaped our own words. Recent work has extended these methods to whale song, finding statistical distributions that closely resemble those found in human language and that appear to arise from learning through cultural transmission rather than from any grammatical rule (Arnon and colleagues, 2025; Youngblood, 2025). This is a good example of how a finding can be both genuinely surprising and easily overstated. 3.3 Biosemiotics and the problem of meaning A third framework asks what a signal means to the animal itself, rather than what it means to the observer. This tradition emphasises that every species inhabits its own perceptual world, and that a signal is only meaningful within that world. An ultrasonic call that we can only hear through equipment is not a marginal signal for a bat, it is the centre of its experience. The practical lesson for research design is that we must be careful about which channel we study. Our own bias toward sound and vision has caused us to under-study chemical and vibrational communication, which are probably the dominant channels on this planet by sheer volume of signalling. 3.4 The framework used in this article Throughout the case studies that follow, three questions will be asked of each system. First, is the signal referential, meaning does it pick out something in the world, or is it purely an expression of the sender's internal state? Second, is the system combinatorial, meaning are units assembled into larger units in a way that changes meaning? Third, is the system learned, meaning does an individual acquire it from others, or is it substantially fixed? These three questions map onto #semantics, #syntax, and cultural transmission, and they are the three axes on which the animal and human cases can be compared most productively. 4. Methods 4.1 Observation and the ethogram The foundation of the field is still careful observation. Researchers construct an ethogram, which is a catalogue of the discrete behaviours a species performs, defined precisely enough that two observers watching the same animal would record the same thing. This is slow, unglamorous work, and it remains irreplaceable, because every later analysis depends on the categories established at this stage. If the units are wrong, everything built on them is wrong. 4.2 Playback experiments The single most important experimental technique is the playback experiment. A recorded signal is played to an animal in a controlled setting, and the response is measured. If a monkey hears an alarm call recorded when a leopard was present, and the monkey runs into a tree, while the same monkey hearing an eagle alarm looks up and hides in a bush, then the calls carry information about the type of predator and not merely about the level of danger. #playback_experiments are powerful because they isolate the signal from its context. Their weakness is that they can only test hypotheses the researcher has already thought of, and they risk habituating the animals if repeated too often. 4.3 Bioacoustic recording and passive monitoring Modern studies rely heavily on automated recorders left in the field for months, on hydrophone arrays in the ocean, and on tags attached to individual animals that record sound, depth, and movement together. This last innovation is critical, because it allows a call to be linked to the individual who produced it and to what that individual was doing at the time. Without individual attribution, a recording of a group is a chorus with no speakers. #passive_acoustic_monitoring produces enormous datasets. A single deployment can generate tens of thousands of hours of audio. This is the point at which computation becomes necessary rather than optional. 4.4 Computational analysis and machine learning The current generation of methods uses deep neural networks to detect, classify, and cluster animal sounds. A detailed review of computational bioacoustics set out both the achievements and the open problems of this approach, including the shortage of labelled training data and the difficulty of generalising a model trained on one population to another (Stowell, 2022). Self-supervised models, which learn useful representations from unlabelled audio, have become a standard tool, since they reduce the dependence on hand-annotated datasets (Hagiwara, 2023). A widely cited perspective article argued that #machine_learning could be used not merely to sort animal sounds but to decode their structure, and it set out the conditions that would need to be met for such a claim to be credible, including the need for well-annotated, context-rich datasets and for experimental validation of any hypothesis generated by a model (Rutz and colleagues, 2023). That last condition deserves emphasis. A model can propose. Only an experiment can confirm. 4.5 The annotation problem Every supervised method depends on labels, and labels depend on human judgement about what counts as a unit and what counts as a context. This is where most hidden assumptions enter the analysis. If a researcher decides in advance that a whale click sequence should be divided at silences longer than two seconds, the resulting "words" are partly an artefact of that decision. Careful papers now report the sensitivity of their results to these choices. Careless papers do not, and students should learn to check. 5. Case Studies in Non-Human Communication 5.1 The honeybee and the problem of displacement The dance of the honeybee remains the strongest and cleanest example of #displacement outside human language. A returning forager performs a figure-of-eight movement on the vertical comb inside a dark hive. The angle of the straight run relative to gravity encodes the direction of a food source relative to the sun. The duration of the run encodes distance. Other bees follow the dancer, then fly out and find the food. The bee is communicating about something that is not present. It is not pointing at the flowers, because the flowers cannot be seen from inside the hive. It is describing a location in a coordinate system. This is displacement, and it is achieved by an animal with roughly one million neurons. For most of the last century the dance was assumed to be entirely innate. A recent experiment challenged that assumption in an elegant way. Colonies of bees were raised without any older experienced dancers, so that the young bees had no models to follow. These bees still danced, but they danced with persistent errors in the encoding of distance, and those errors were never fully corrected even after they gained their own foraging experience. Bees raised in normal colonies, with older dancers to observe, performed accurately (Dong, Lin, Nieh, and Tan, 2023). The implication is significant. The #honeybee_waggle_dance has a learned component. Social learning refines an inherited behaviour, and a colony can pass a local dialect of distance encoding across generations. This is a form of #cultural_transmission in an insect. What the dance lacks is equally instructive. It is not compositional. There are no units that recombine to produce new meanings. It communicates about one topic, food location, and cannot be repurposed to discuss anything else. It is a specialised instrument, not a general one. 5.2 Alarm calls and functional reference The classic work on vervet monkeys established that these animals produce acoustically distinct #alarm_calls for leopards, eagles, and snakes, and that listeners respond with the appropriate escape behaviour even when the predator is absent and the call is played from a hidden speaker. This finding shaped the field for forty years and gave rise to the concept of functional reference: a signal that behaves as though it refers to an external object, whether or not the sender has any internal concept of that object. The concept has since been criticised. The alternative explanation is that the calls encode the appropriate response, or the sender's own state, rather than the predator itself. A leopard call might mean "climb" rather than "leopard". Both accounts predict the same behaviour, which is precisely why the distinction is difficult to test. A useful way through this problem is to test whether the signal can be used flexibly across contexts. If the same call appears in a situation with no predator but with a similar required response, the response-based account gains support. If the call is tightly bound to one class of object regardless of the response, the referential account gains support. Recent primate research has moved toward more careful, more sceptical interpretations, emphasising that primate vocal production is largely inflexible while primate comprehension is remarkably flexible, an asymmetry that may be the key to understanding why our own lineage broke the pattern (Fischer, 2021). #urgency_coding is a further complication. In many species, calls vary continuously with how frightened the animal is, and much of the apparent information content can be explained by arousal alone. Good studies control for this by matching the intensity of the stimulus across conditions. 5.3 Compositionality in birds The most rigorous demonstration of #compositionality in a wild animal comes from work on the Japanese tit. These small birds produce an alert call and a recruitment call. Played alone, the alert call causes scanning behaviour. Played alone, the recruitment call causes approach. Played together in the natural order, the birds both scan and approach, which is the behaviour used to mob a predator. Played in the reversed order, which does not occur in nature, the birds show a much weaker response. This shows that the meaning of the combination depends on the order of the parts, which is the definition of a compositional rule. Further experiments have pushed this further, testing whether the birds treat the combination as a single unit built from two constituents rather than as two independent messages arriving in sequence. Evidence has been reported that the birds form an integrated representation from the two calls, consistent with a basic merge-like operation (Suzuki and Matsumoto, 2022). This is a striking result. A songbird with a brain the size of a pea appears to perform a rudimentary form of the combinatorial operation that syntacticians place at the centre of human grammar. The correct conclusion is not that the tit has grammar. It is that the combinatorial operation is cheap, and available to small brains, and therefore cannot be the thing that made human language unique. #birdsong more generally offers another lesson. Songbirds are one of the few groups outside humans with true #vocal_learning. A young bird raised in isolation sings an abnormal song. A young bird exposed to a tutor sings the tutor's song. Populations develop dialects. This is a genuine parallel with human speech acquisition, and it is why songbirds are a central model system for the neuroscience of learned vocalisation. But song is used almost entirely for territory and mating. It is elaborate in form and narrow in function, which is close to the opposite of human language. 5.4 Great ape gestures The ape case is the one most often distorted in public discussion, and it deserves careful handling. Wild great apes use a large repertoire of gestures, on the order of eighty or more distinct forms, produced with the hands, the limbs, and the whole body. These gestures are used intentionally in the technical sense: the ape checks whether the recipient is looking, waits for a response, and persists or elaborates if the response does not come. Those three criteria, audience checking, response waiting, and persistence, are the standard behavioural markers of intentional signalling, and apes meet them. An unusual and revealing study tested whether humans with no experience of apes could understand ape gestures. Participants watched video clips of #chimpanzee_gestures and bonobo gestures and were asked to select the intended meaning from a set of options. They performed well above chance, even without training (Graham and Hobaiter, 2023). This suggests that some of the ape gestural repertoire is shared with humans, and that the meanings are recoverable by an untrained human observer. That is a remarkable result, and it points to a deep continuity in the #primate_gestures system. It also has clear limits. The gestures are used to request, to invite, to stop, to move. They are imperative rather than declarative. Apes ask for things. They very rarely comment on things for the sake of sharing attention, which human infants do constantly and spontaneously from around the first year of life. 5.5 The ape language projects and what they taught us Between the 1960s and the 1990s, several projects attempted to teach human language to apes, using sign language or symbol boards, since apes lack the vocal tract control for speech. The results were widely publicised and are still widely misremembered. The apes learned symbols. Some learned several hundred. They used symbols to request objects and activities, and some produced novel combinations. This is not nothing, and it should not be dismissed. But careful reanalysis of the data revealed severe problems. The proportion of utterances that were prompted or partially imitated from the trainer's immediately preceding signs was very high. Utterance length did not grow with age, unlike in human children, whose mean utterance length increases steadily. Combinations were often repetitive rather than structured. And the recording practices in several projects were not blind, which opens the door to the #Clever_Hans_effect, in which an animal responds to unconscious cues from the human rather than to the task itself. The lesson for the modern field is methodological rather than dismissive. Where an experimenter knows the expected answer and is present during the trial, the experiment is compromised. Blind coding, double-blind design, and pre-registration of predictions are now standard for exactly this reason. Students who read the older literature should read it with these problems in mind, and should be careful about citing findings that would not survive a modern protocol. 5.6 Sperm whales and combinatorial coding Sperm whales communicate using codas, which are short patterned sequences of clicks. Different clans in the Caribbean and the Pacific use different coda repertoires, and clan membership is defined by these repertoires rather than by geography or genetics. This is one of the clearest cases of #cultural_transmission in a marine mammal. Recent computational work on a large annotated dataset from a long-term study population has changed the picture substantially. By analysing codas in their conversational context rather than in isolation, researchers identified systematic variation along several independent dimensions, including the rhythm of the clicks, the overall tempo, an extension of duration across an exchange that they called rubato, and the addition of an extra click that they called an ornament. These features vary in a graded and context-sensitive way, and they combine to yield a much larger inventory of distinguishable coda types than had previously been recognised (Sharma and colleagues, 2024). The authors described this as a combinatorial coding system, and drew a cautious comparison to the way human phonetic systems build a large inventory of forms from a small set of independently varying features. Earlier work had already set out the roadmap for this kind of analysis and the interdisciplinary effort required to pursue it (Andreas and colleagues, 2022). What remains unknown, and this must be stated plainly, is what any of it means. The structure is real. The mapping from structure to meaning has not been established. #sperm_whale_codas may turn out to encode identity, coordination, emotional state, or something we have not considered. Structure is a necessary condition for a language-like system, but it is not a sufficient one. #Project_CETI, the large collaborative effort working on this species, is explicit about this distinction, and its own publications are more careful than most of the reporting about them. 5.7 Humpback song and statistical structure Male humpback whales produce long, hierarchically organised songs that change over time. All the males in an ocean basin sing approximately the same song at the same time, and the song evolves through the season. Occasionally a new song sweeps across a population, replacing the old one within a year or two. This is population-level cultural change of a kind rarely documented in any animal. A recent analysis applied methods from the study of infant language acquisition to decades of #humpback_song recordings. Human infants segment continuous speech into words partly by tracking transitional probabilities between syllables, noticing that some transitions are much more predictable than others. When the same segmentation procedure was applied to whale song, the resulting units followed a Zipfian frequency distribution, in which a few units are very common and a long tail is very rare, closely matching the distribution found in human language (Arnon and colleagues, 2025). Complementary work found evidence of efficiency effects consistent with linguistic laws (Youngblood, 2025). The interpretation offered by the authors is important and is frequently lost in coverage of this work. They do not claim that whale song is language, and they do not claim it carries propositional meaning. They argue that this statistical structure is a signature of learning under cultural transmission. Any system that is passed from individual to individual by imitation, and that is under pressure to be learnable, will tend to develop this kind of skewed structure. The finding therefore tells us about the transmission process, not about the semantics. This is a good example for students of how to read a strong result correctly. The finding is real and important. It supports a claim about learning. It does not support a claim about meaning, and the authors did not make one. 5.8 Dolphins and individual identity Bottlenose dolphins develop #signature_whistles, which are individually distinctive whistle contours that a dolphin invents in its first year and keeps for life. Dolphins use their own signature whistle to broadcast identity, and they can copy the signature whistle of a close associate, apparently to address that individual. When a dolphin hears a copy of its own signature whistle played back, it responds. This is important because it demonstrates something close to a label for an individual, which is a form of reference to a specific entity rather than to a category. It also requires vocal production learning, since the whistle is invented rather than inherited. Recent efforts have applied large model architectures to dolphin vocalisations, aiming to identify recurrent structure and eventually to generate synthetic whistles that dolphins might respond to. These projects are at an early stage. The honest summary is that we can now detect and cluster dolphin sounds with high reliability, and we cannot yet say what most of them mean. 5.9 Elephants and name-like calls A study of wild African savannah elephants tested whether the low frequency rumbles these animals use over long distances contain individually specific components addressed to particular receivers. Machine learning models were trained to predict the intended receiver from the acoustic structure of a call, and performed above chance. In playback tests, elephants responded more strongly, and more quickly, to recordings of calls that had originally been addressed to them than to calls that had been addressed to a different individual (Pardo and colleagues, 2024). This is evidence for something functionally similar to a name. Crucially, it is not vocal imitation. The dolphin case works by copying the other individual's own signature. The elephant case appears to work by using an arbitrary label that does not imitate the receiver. If this holds up under replication, it is a closer analogue to a human name than anything else documented in a non-human animal, because it combines reference to an individual with #arbitrariness. #elephant_calls also illustrate the value of long-term field sites. This result was only possible because researchers had decades of records identifying individual elephants and their relationships. 5.10 Marmosets and vocal labels In the same year, a study of common marmosets found that these small monkeys use specific calls, known as phee calls, that appear to encode the identity of the individual being addressed. Marmosets responded more strongly to calls directed at them. Members of the same family group used similar labels for the same individual, and even unrelated adults living together converged on shared labelling conventions (Oren and colleagues, 2024). The convergence result is the striking part. It suggests learned, group-specific conventions for referring to individuals, which is a hallmark of a socially transmitted system rather than a genetically fixed one. #marmoset_names, like elephant names, involve using a sound to pick out a specific individual who is not the sender. Two independent findings of this kind, in two very distantly related mammals, in the same year, suggest that the capacity to label individuals may be more widespread than assumed, and that it may evolve wherever animals live in groups that are large enough and stable enough for individual recognition to matter. 5.11 Bats, rodents, and the unheard world Bats are one of the very few mammal groups with demonstrated vocal production learning, and they have become an important model system for exactly that reason. Some species produce complex vocal exchanges in roosts, including calls that appear to be directed at specific individuals and that vary with social context. A broad review of #vocal_learning across taxa argued that the trait is not a single capacity but a multi-dimensional cluster of abilities, which may appear in different combinations in different lineages (Vernes and colleagues, 2021). This reframing is useful, because it explains why species seem to fall on a continuum rather than into two clean categories. #bat_communication also underlines the perceptual point made earlier. Much of it is ultrasonic. It was effectively invisible to science until the technology existed to record it. #prairie_dogs have been the subject of long-running claims that their alarm calls encode the size, colour, and species of an approaching predator. These claims are interesting, they have not been widely replicated by independent groups, and they remain contested. Students should treat them as an open question rather than as an established finding, and should note the general principle that extraordinary claims in this field require independent replication. 5.12 Beyond animals: acoustic emission in plants A finding that widened the scope of the field concerned plants. Tomato and tobacco plants under stress, from drought or from physical damage, emit airborne ultrasonic clicks. These sounds are within the hearing range of many insects and mammals. Machine learning classifiers trained on the recordings could distinguish a dry plant from a cut plant from an unstressed plant with high accuracy (Khait and colleagues, 2023). This is not communication in the biological sense, at least not yet. The authors were careful about this. There is no evidence that the plant produces the sound in order to influence a receiver, and the sounds may simply be a by-product of cavitation in the water-conducting tissue. But the sounds are informative, they are detectable, and they could in principle be exploited by other organisms, which is how many communication systems begin evolutionarily. #plant_acoustics is now a serious research area, and it is a reminder that our category boundaries are provisional. 5.13 Cephalopods and the visual channel #cephalopods deserve mention because they communicate largely in a channel we can see but do not use. Cuttlefish, squid, and octopuses can change the colour, pattern, and texture of their skin within a fraction of a second, using chromatophores under direct neural control. Some squid produce patterned body displays during interactions that appear to have distinct components, and some species can even send different displays from the two sides of the body simultaneously, one to a female and one to a rival. The bandwidth of this system is extremely high. Its structure is poorly understood, in part because the units are hard to define and in part because these animals are difficult to observe in the wild for long periods. It is one of the most promising and least explored areas in the field. 5.14 Domestic animals and the interspecies channel Dogs represent a special case, because they have evolved alongside humans and are unusually good at reading human signals, including pointing and gaze, which even chimpanzees find difficult. Dogs learn labels for objects, and some individuals learn a large number of them. The recent enthusiasm for #dog_soundboards, in which dogs press buttons that play recorded words, has produced viral videos and a small but growing body of published research. The scientific questions are whether button presses are non-random, whether they are context-appropriate, and whether combinations carry more than the sum of their parts. Early evidence suggests that presses are indeed non-random and often context-appropriate. Whether the dogs are combining concepts, or have learned that a particular sequence of presses reliably produces a particular outcome, is not yet settled, and the field will need careful blind testing to settle it. The history of the ape projects is directly relevant here, and the same methodological cautions apply. #interspecies_communication of this kind is scientifically interesting and also socially significant, because it shapes public beliefs about animal minds far more powerfully than any journal article. 6. What the Evidence Shows, Property by Property Having surveyed the cases, we can now return to the design features and score them. Reference. Present in a functional sense in many species. Alarm calls, name-like calls, and food calls all pick out categories or individuals in the world. Whether the animal holds a concept corresponding to the referent is not established, and may not be testable with current methods. Arbitrariness. Present in some cases. Elephant name-like calls, if confirmed, are arbitrary labels. A theoretical framework has been proposed for treating arbitrariness not as a binary property but as a matter of degree, with signals distributed along a continuum from strongly motivated to fully arbitrary (Watson and colleagues, 2022). This is a more useful way to compare species than a yes or no test. Displacement. Present in the honeybee, and very rare elsewhere. This is one of the most surprising distributions in the whole field. The capacity appears in an insect and is largely absent in our closest relatives. Chimpanzee gestures are almost entirely about the here and now. Compositionality. Present in at least one wild bird and probably in several primates. The evidence from Japanese tits is the strongest. A thorough critical review of the methods used to test for animal syntax and semantics set out the standards that such claims must meet, and argued that many published claims do not meet them (Berthet, Coye, Dezecache, and Kuhn, 2023). Students should read this kind of review before accepting any specific claim about animal grammar. Duality of patterning. Possibly emerging in sperm whale codas, if the independent feature dimensions identified in recent work function as meaningless units that combine into meaningful ones. This remains a hypothesis. Cultural transmission. Widespread. Present in songbirds, whales, parrots, bats, and now in honeybees. This is one of the clearest continuities between human and non-human systems. Open-ended productivity. Not demonstrated anywhere. This is the crucial gap. Every known animal system has a bounded repertoire. Animals do not appear to produce or understand an unlimited set of novel messages. Some combinatorial flexibility exists, but nothing approaching the generative capacity that a four year old human child possesses. Recursion. Not demonstrated. Claims have been made and have not survived scrutiny. #turn_taking. Present in many species, including marmosets, some birds, and elephants. Conversational timing, with gaps of a few hundred milliseconds and avoidance of overlap, appears to be an ancient feature that predates language. The overall picture is of a mosaic. Individual components of language are scattered across the animal kingdom, appearing in different lineages for different reasons. What is unique about humans is not any single component. It is the assembly of all of them into one system, together with an open-ended productivity that no other species has shown. 7. Computation, Decoding, and the Limits of Pattern Finding 7.1 What the new tools can do Deep learning has already transformed the practical side of the field. Models can detect calls in noisy recordings, separate overlapping sources, assign calls to individuals, and cluster a repertoire into candidate types without human labels. These are enormous gains in efficiency, and they make possible studies that could not have been attempted a decade ago. #deep_learning has also enabled large-scale monitoring for #conservation, where models trained on audio can count animals, track populations, and detect illegal activity across landscapes too large to survey by other means (Tuia and colleagues, 2022). Neuroethological approaches now combine bioacoustic analysis with recordings from the brain, allowing researchers to ask how the units identified by a computational model correspond to units represented in the animal's nervous system (Sainburg and Gentner, 2021). This is a powerful validation strategy, because it grounds the analysis in the animal's own processing rather than in the researcher's segmentation choices. 7.2 What the new tools cannot do The core limitation is that meaning is not recoverable from form alone. A large model trained on whale sound will learn the statistical structure of whale sound. It will not learn what whales are talking about unless the training data include information about the context in which each sound was produced. This is why the most important bottleneck is not compute. It is annotated, context-rich data: recordings in which we know who called, when, to whom, in what social situation, and what happened next. Collecting this requires long-term field sites, individual identification, and years of patient work. The computational revolution has, if anything, increased the value of traditional fieldwork rather than replacing it. 7.3 The translation fallacy Public discussion often assumes that decoding animal communication will work like machine translation between two human languages. It will not, for a structural reason. Machine translation works because the two languages describe overlapping worlds. A French speaker and an English speaker both have concepts for bread, for anger, for tomorrow, and this shared conceptual space is what allows alignment between the two vocabularies. A sperm whale's world overlaps with ours much less. It has no hands, no fire, no writing, and a sensory system built around sound in three dimensions at depths we cannot survive. If sperm whales do have something like words, many of those words may have no counterpart in any human language, and no amount of statistical alignment will produce one. The realistic goal is not translation. It is the identification of structure, followed by experimental testing of hypotheses about function, followed slowly and partially by an understanding of what a small number of signals do. That is a worthwhile goal, and it is achievable. It is not the same as talking to a whale. 7.4 The risk of playback with generated signals Several projects intend to synthesise signals and play them back to animals to test whether the animals respond as predicted. This is a legitimate and powerful experimental design. It also raises a serious ethical issue. Broadcasting artificial signals into a wild population could disrupt social behaviour, cause stress, or interfere with foraging and reproduction, particularly in endangered species. #research_ethics in this area is not a formality. Any playback of generated signals should be preceded by a risk assessment, and should be conducted at the smallest scale that can answer the question. 8. Critical Perspectives 8.1 Anthropomorphism and its opposite Anthropomorphism is the attribution of human characteristics to animals without evidence. It is a real and persistent bias. Humans are extremely good at seeing intention, and we see it whether or not it is there. The opposite error is less discussed but equally damaging. It has been called anthropodenial: the reflexive refusal to attribute any human-like capacity to an animal, even when the evidence supports it. A researcher who assumes in advance that an elephant cannot have a name will design a study that cannot detect one. The solution is not to pick a side but to make the hypotheses explicit and testable. If two accounts predict the same behaviour, the experiment is not decisive, and the honest response is to design a better experiment rather than to assert a preference. 8.2 The statistical trap With enough data and enough analytical freedom, structure can be found in almost anything. Randomly generated sequences can be made to look Zipfian with the right unit definitions. This is why pre-registration matters, why null models matter, and why a result should be tested against surrogate data generated by shuffling the original in ways that destroy the hypothesised structure while preserving everything else. Students evaluating a paper in this area should ask three questions. How were the units defined, and would a different definition change the result? What null model was used? Was the key prediction stated before the data were analysed? 8.3 Publication and media incentives Findings in this field attract exceptional media attention, which creates pressure toward strong claims. The gap between what a paper says and what a headline says is frequently large. Careful authors write that a structure is "consistent with" a combinatorial system. The headline says that whales have an alphabet. Both the scientists and the journalists share responsibility here, and the cost is borne by public understanding. 8.4 The species bias in the literature The literature is dominated by a small number of charismatic species: chimpanzees, dolphins, elephants, and a handful of songbirds. This reflects funding, public interest, and the practicalities of observation. It also means that our theories of #signal_evolution are built on an unrepresentative sample. Most communication on Earth is chemical, is produced by invertebrates, and is almost entirely unstudied. Any general theory built on primates and cetaceans should be regarded as provisional. 9. Why This Matters 9.1 For the evolution of language The comparative method is the only empirical route into the evolution of human language, because language does not fossilise. By identifying which components exist in which species, we can reconstruct which capacities were present in our ancestors and which are recent. The pattern emerging from recent work suggests that most of the individual components of language are old and widely shared. Reference, combination, turn-taking, cultural transmission, and vocal learning all exist elsewhere. What appears to be genuinely rare is the combination of these components into a single unbounded system, along with the motivation to share information for its own sake. If that is correct, then the study of #language_origins should focus less on the emergence of syntax and more on the emergence of the social motivation to inform others, which apes largely lack and human infants show from the beginning. 9.2 For conservation Acoustic monitoring is now a standard conservation tool. It can identify species presence without capture, detect population declines early, and measure the impact of noise pollution. Understanding what animals are communicating makes these tools sharper. If a particular call type indicates stress, then the rate of that call becomes a welfare and habitat indicator that can be measured automatically at scale. Noise pollution is itself a communication problem. Shipping noise masks whale calls and reduces the distance over which they can coordinate. Urban noise causes birds to sing at higher pitches, which may reduce their attractiveness to mates. These are direct, measurable harms that follow from taking animal communication seriously. 9.3 For welfare and ethics If an animal can label individuals, coordinate through learned conventions, and pass a communication system down through generations, then its social and cognitive life is richer than most policy frameworks assume. #animal_welfare regulation has historically focused on the capacity to feel pain. The evidence reviewed here concerns a different capacity: the capacity to participate in a social world made partly of learned signals. This does not resolve any ethical question by itself. Facts about capacity do not automatically generate obligations. But they do change what the ethical question is about. Arguments about animal moral status that rest on assumptions of cognitive simplicity are weakened when the empirical basis for those assumptions is removed. 9.4 For our understanding of ourselves Perhaps the deepest reason to study #animal_cognition and communication is that it provides the only external reference point we have for understanding human language. Without comparison, we cannot know which features of our language are necessary consequences of any communication system, which are accidents of our lineage, and which are genuinely unusual. #comparative_linguistics across species is the only way to answer that, and it is a young field with most of its work still ahead. 10. Open Questions and a Research Agenda The following questions are unresolved and are suitable starting points for student research projects. Unit definition. How can the units of an animal signal system be identified without importing human assumptions? Can unsupervised methods and neural recordings converge on the same segmentation? #data_annotation choices remain the single largest source of hidden bias in the field. Meaning. What experimental designs can distinguish a signal that refers to an object from a signal that specifies a response? This is the central unsolved problem in animal #semantics. #pragmatics. Almost nothing is known about how animals use signals strategically, how they handle ambiguity, or whether they adjust signals for the knowledge state of the receiver. Some evidence of audience effects exists, and it is thin. #context_dependence. How much of the apparent complexity in animal repertoires is graded variation driven by arousal, and how much is discrete and categorical? This question can be attacked directly with careful playback and with modern classification methods. Chemical and vibrational channels. These are the largest unexplored territories. The methods exist. The attention does not. Replication. Several high-profile claims in this field have never been independently replicated. Replication studies are undervalued and badly needed, and they are well suited to student projects because they require rigour rather than novelty. #honest_signals and deception. Under what conditions do animals produce misleading signals, and how do receivers guard against them? This connects the field to evolutionary game theory and remains under-tested in the wild. 11. Conclusion Animals do not have language, if by language we mean the open-ended, recursive, socially motivated system that a human child acquires without instruction. That conclusion has survived a century of attempts to overturn it, and the failure of the twentieth century ape projects to overturn it was instructive rather than disappointing. But this conclusion, stated on its own, is misleading, because it obscures what has actually been discovered. Bees describe places they cannot see. Birds combine calls in an order that changes meaning. Whales pass songs across oceans and shape them into distributions that look like the distributions in our own speech. Elephants and marmosets appear to address one another with learned, arbitrary labels. Apes gesture with intent, and we understand them without training. None of this is language. All of it is more than we believed. The most defensible position, and the one supported by the evidence reviewed here, is that language is not one thing that we have and animals lack. It is a bundle of capacities, and almost every strand in the bundle can be found somewhere else in nature. Our species did not invent the strands. We tied them together. For students entering this field, the work ahead is not to prove that animals talk. It is to describe, precisely and sceptically, what they actually do, and to build the methods that will let the next generation ask better questions than we can. That is slower than the headlines suggest, and it is more interesting. Hashtags #Animal_Languages #Do_Animals_Have_Language #Animal_Communication_Research #Talking_Animals #Whale_Language #Bee_Dance_Language #Bird_Call_Syntax #Ape_Sign_Language #Decoding_Animal_Sounds #Language_Evolution_Science #Nonhuman_Minds #Speaking_With_Whales #Bioacoustics_Research #Animals_And_Meaning #Science_Of_Listening References Andreas, J., Begus, G., Bronstein, M. M., Diamant, R., Delaney, D., Gero, S., Goldwasser, S., Gruber, D. F., de Haas, S., Malkin, P., Payne, R., Petri, G., Rus, D., Sharma, P., Tchernov, D., Tonjes, P., Zacarian, K., and Zitterbart, D. P. (2022). Toward understanding the communication in sperm whales. iScience. Arnon, I., Kirby, S., Allen, J. A., Garrigue, C., Carroll, E. L., and Garland, E. C. (2025). Whale song shows language-like statistical structure. Science. Berthet, M., Coye, C., Dezecache, G., and Kuhn, J. (2023). Animal linguistics: a primer. Biological Reviews. Bakker, K. (2022). The Sounds of Life: How Digital Technology Is Bringing Us Closer to the Worlds of Animals and Plants. Princeton University Press. Dong, S., Lin, T., Nieh, J. C., and Tan, K. (2023). Social signal learning of the waggle dance in honey bees. Science. Fischer, J. (2021). Primate vocal communication and the evolution of speech. Current Directions in Psychological Science. Graham, K. E., and Hobaiter, C. (2023). Towards a great ape dictionary: inexperienced humans understand common nonhuman ape gestures. PLOS Biology. Hagiwara, M. (2023). AVES: animal vocalization encoder based on self-supervision. Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing. Kershenbaum, A. (2024). Why Animals Talk: The New Science of Animal Communication. Viking, Penguin Random House. Khait, I., Lewin-Epstein, O., Sharon, R., Saban, K., Goldstein, R., Anikster, Y., Zeron, Y., Agassy, C., Nizan, S., Sharabi, G., Perelman, R., Boonman, A., Sade, N., Yovel, Y., and Hadany, L. (2023). Sounds emitted by plants under stress are airborne and informative. Cell. Mustill, T. (2022). How to Speak Whale: A Voyage into the Future of Animal Communication. Grand Central Publishing. Oren, G., Shapira, A., Lifshitz, R., Vinepinsky, E., Cohen, R., Fried, T., Hadad, G. P., and Omer, D. (2024). Vocal labeling of others by nonhuman primates. Science. Pardo, M. A., Fristrup, K., Lolchuragi, D. S., Poole, J. H., Granli, P., Moss, C., Douglas-Hamilton, I., and Wittemyer, G. (2024). African elephants address one another with individually specific name-like calls. Nature Ecology and Evolution. Rutz, C., Bronstein, M., Raskin, A., Vernes, S. C., Zacarian, K., and Blasi, D. E. (2023). Using machine learning to decode animal communication. Science. Sainburg, T., and Gentner, T. Q. (2021). Toward a computational neuroethology of vocal communication: from bioacoustics to neurophysiology, emerging tools and future directions. Frontiers in Behavioral Neuroscience. Semple, S., Ferrer-i-Cancho, R., and Gustison, M. L. (2022). Linguistic laws in biology. Trends in Ecology and Evolution. Sharma, P., Gero, S., Payne, R., Gruber, D. F., Rus, D., Torralba, A., and Andreas, J. (2024). Contextual and combinatorial structure in sperm whale vocalisations. Nature Communications. Stowell, D. (2022). Computational bioacoustics with deep learning: a review and roadmap. PeerJ. Suzuki, T. N., and Matsumoto, Y. K. (2022). Experimental evidence for core-Merge in the vocal communication system of a wild passerine. Nature Communications. Tuia, D., Kellenberger, B., Beery, S., Costelloe, B. R., Zuffi, S., Risse, B., Mathis, A., Mathis, M. W., van Langevelde, F., Burghardt, T., Kays, R., Klinck, H., Wikelski, M., Couzin, I. D., van Horn, G., Crofoot, M. C., Stewart, C. V., and Berger-Wolf, T. (2022). Perspectives in machine learning for wildlife conservation. Nature Communications. Vernes, S. C., Kriengwatana, B. P., Beeck, V. C., Fischer, J., Tyack, P. L., ten Cate, C., and Janik, V. M. (2021). The multi-dimensional nature of vocal learning. Philosophical Transactions of the Royal Society B: Biological Sciences. Watson, S. K., Filippi, P., Gasparri, L., Falk, N., Tamer, N., Widmer, P., Manser, M., and Glock, H. J. (2022). Optionality in animal communication: a novel framework for examining the evolution of arbitrariness. Biological Reviews. Youngblood, M. (2025). Language-like efficiency in whale communication. Science Advances.

International commerce rests on a simple promise: parties who agree to a deal are expected to keep it. Yet the last few years have shown how quickly that promise can be tested. A pandemic can shut borders overnight, a war can turn trade routes into conflict zones, and a single blocked shipping lane can send shocks through supply chains on the other side of the planet. When these shocks arrive, two legal ideas step forward to decide who bears the loss: #force_majeure and #hardship. This article explains the difference between the two, traces where they came from, and compares how they operate across common law systems, civil law systems, the United Nations Convention on Contracts for the International Sale of Goods, the UNIDROIT Principles, and the model clauses of the International Chamber of Commerce. It then stress-tests both doctrines against three modern crises: the COVID-19 #pandemic, the war in Ukraine and the #sanctions that followed, and large-scale supply-chain failures. The central argument is that force majeure and hardship answer two different questions. Force majeure asks whether performance has become impossible or seriously obstructed, and it usually leads to suspension or release. Hardship asks whether performance has become so unbalanced that it is unfair to demand it unchanged, and it usually leads to #renegotiation or judicial adaptation. Confusing the two is one of the most common and costly drafting errors in cross-border trade. The article concludes with practical guidance for students, drafters, and future practitioners, and it identifies open questions for further research. Keywords: force majeure; hardship; international commercial contracts; CISG Article 79; UNIDROIT Principles; risk allocation 1. Introduction Every contract is, in a sense, a bet on the future. Parties look at the world as it is on the day they sign, and they assume the world will stay close enough to that picture for the deal to make sense. Most of the time that assumption holds. Goods are shipped, invoices are paid, and the contract quietly does its job. The interesting legal questions begin when the world refuses to cooperate, when the ground shifts so far under a contract that keeping the promise becomes impossible, dangerous, or financially crushing. Two doctrines govern that moment. The first is #force_majeure, a French term meaning superior force, which addresses events that block performance entirely or make it impossible to carry out. The second is #hardship, which addresses events that leave performance technically possible but so unbalanced that demanding it unchanged would produce a kind of #economic_ruin for one side. These two concepts are cousins, they often appear in the same contract, and they are constantly mistaken for each other. That confusion matters. A party that pleads force majeure when it should have pleaded hardship, or the reverse, will often lose. The stakes have risen sharply in recent years. The COVID-19 pandemic produced a wave of disputes in which companies argued that lockdowns, border closures, and factory shutdowns excused them from their obligations (Natarajan, 2022). Soon after, the war in Ukraine and the layers of #sanctions that governments imposed created another surge of claims, this time centred on whether legal restrictions and physical danger fell within the meaning of a force majeure clause (Zeller, 2022). Alongside both crises ran a quieter but equally serious problem: the breakdown of global #supply_chain networks, where a shortage in one country made it impossible or ruinous for a supplier three borders away to deliver on time. This article has four goals. First, it aims to give students a clear and honest account of what force majeure and hardship actually mean, stripped of the jargon that often surrounds them. Second, it compares how different legal systems and international instruments treat these ideas, because a contract governed by English law behaves very differently from one governed by German or French law, and both behave differently again from one governed by the CISG. Third, it tests the doctrines against real crises, showing where they help and where they fail. Fourth, it offers concrete drafting and negotiation advice, because the single most important lesson of the last few years is that the words parties choose in advance matter far more than the doctrines they invoke afterward. The argument throughout is that these doctrines are tools of #risk_allocation. They decide, when the unexpected strikes, who absorbs the loss. Understanding them well is not an academic luxury. It is the difference between a business that survives a shock and one that does not. The rest of the article proceeds as follows. Section 2 reviews the recent literature. Section 3 sets out the method. Sections 4 and 5 build the conceptual and historical foundations. Section 6 compares the main international instruments. Section 7 tests the doctrines against three modern crises. Section 8 discusses drafting and the deeper choice between excuse and adaptation. Section 9 offers practical guidance. Section 10 notes the limits of the study and points to future research, and Section 11 concludes. 2. Literature Review Scholarship on excused non-performance is old, but the recent literature has been reshaped almost entirely by the events of the 2020s. Before the pandemic, force majeure clauses were often treated as routine boilerplate, the fine print that nobody read closely. Writers such as Palmer (2022) have since traced how the classic doctrines of #impossibility, #impracticability, and #frustration grew out of a much stricter starting point, in which contracts were treated as nearly absolute and excuses were rare. Palmer shows that modern law slowly softened this position, carving out narrow escapes for events that no reasonable party could have controlled or foreseen, while still keeping the core principle that promises should be kept. A large body of writing has grown specifically around the CISG and its central excuse provision, Article 79. Janssen and Wahnschaffe (2021) examined whether the pandemic offered genuinely new grounds for exemption or whether existing rules already covered the situation. Their conclusion, broadly, was that the framework was largely adequate but that many disputes turned on precise timing and #foreseeability rather than on the raw fact of a global crisis. Jevremovic (2022) tested the same provision through the lens of the outbreak and asked whether Article 79 truly protects international trade when performance becomes impossible, concluding that the provision works reasonably well for physical impediments but remains uncertain when the problem is purely economic. Zeller (2022) revisited Article 79 in light of the pandemic and reinforced the point that courts and tribunals have been reluctant to treat rising costs alone as a valid excuse. Several writers have focused on the harder question of hardship within instruments that do not clearly provide for it. Perales Viscasillas (2021) analysed the doctrine of changed circumstances, known in the civil law tradition as #rebus_sic_stantibus, within the CISG and various soft-law instruments, highlighting the ongoing debate over whether the Convention leaves any room for relief when a deal becomes badly unbalanced. Qiu, Xu, and Yuan (2023) studied the meaning of the word impediment in Article 79 and compared how the concept relates to both force majeure and hardship, drawing particular attention to how Chinese courts have handled pandemic claims. Their work is useful because it shows a real legal system wrestling in real time with the boundary between the two doctrines. Other scholars have looked at the practical machinery of dispute resolution. Kiraz and Yildiz Ustun (2020) reviewed arbitral awards dealing with force majeure clauses and found that tribunals paid close attention to the exact wording of each clause rather than to any universal definition. Bix (2022) offered a more sceptical view, arguing that equitable relief during the pandemic should be approached with realism rather than optimism, since courts are cautious about rewriting bargains. Rothstein (2022) examined the strategies parties used to try to escape performance during the pandemic and warned that many of these arguments were weaker than the parties assumed. Guimaraes and colleagues (2021) studied the contractual performance crisis in international sales governed by the CISG and confirmed that the pandemic exposed both the strengths and the gaps of the existing rules. Taken together, the recent literature agrees on a few key points. Physical impossibility is treated far more generously than economic difficulty. The wording of a clause almost always beats the abstract doctrine behind it. And the line between force majeure and hardship, though blurry in practice, remains legally important. Where the literature still divides is on the harder questions: whether purely economic hardship should ever excuse performance, how far a decision-maker should go in adapting a contract, and how long an ongoing crisis can be treated as extraordinary before it becomes a normal business risk. This article builds on the points of agreement and engages directly with the points of division, extending the discussion by comparing the doctrines side by side across systems and by testing them against three distinct modern crises. 3. Methodology This article uses a comparative doctrinal method, which is the standard approach for legal research of this kind. Rather than collecting numerical data, it examines legal rules, principles, and instruments to understand how they are structured and how they interact. The method has three steps. The first step is conceptual. The article defines force majeure and hardship carefully and separates them, because much of the confusion in practice comes from treating them as a single idea. The second step is comparative. The article places several legal systems and international instruments next to each other, including the common law tradition, the civil law tradition, the CISG, the UNIDROIT Principles of International Commercial Contracts, and the ICC model clauses. Comparison reveals not only differences in wording but differences in underlying philosophy, especially around the question of whether a court should ever adjust a contract that has become unbalanced. The third step is applied. The article takes three real crises, the pandemic, the war and sanctions in Ukraine, and supply-chain failures, and asks how each doctrine performs when tested against them. The sources used are peer-reviewed journal articles and academic book chapters published within the last few years, chosen because they reflect the most current thinking after the shocks of the 2020s. Selecting recent sources matters here, because the practical understanding of these doctrines has shifted more in the last five years than in the previous fifty. The analysis does not attempt to resolve every disputed point of national law, which would be impossible in a single article. Instead it aims to give a clear map of how the doctrines relate to one another and how a careful drafter or negotiator can use that map. The intended readers are students and early practitioners, so the writing favours clarity over technical density, while keeping the structure and rigour expected of a formal research article. 4. Conceptual Distinctions: What Force Majeure and Hardship Actually Mean 4.1 The Meaning of Force Majeure #force_majeure is the older and more intuitive of the two ideas. Its core question is whether an event beyond the control of the parties has made performance impossible, or has so seriously obstructed it that carrying on is no longer realistic. The traditional examples are the ones people picture when they hear the phrase #acts_of_God: earthquakes, floods, storms, and other natural disasters. Over time the category expanded to include human events of similar force, such as war, riot, government prohibition, and general strikes. Three features usually define a genuine force majeure event. It must be external, meaning it comes from outside the control of the affected party. It must be unforeseeable, or at least not reasonably foreseeable at the time the contract was made. And it must be unavoidable, meaning the affected party could not have prevented or overcome it through reasonable effort. If all three features are present, and if the event actually prevents performance rather than merely making it harder, the affected party is usually excused. The typical result is that the obligation is suspended for as long as the event lasts, and if the event lasts long enough, the contract may be terminated. A crucial and often misunderstood point is that force majeure is about prevention, not expense. A supplier whose factory has been destroyed by an earthquake has a strong claim. A supplier whose costs have tripled because raw material prices spiked does not, at least not under classic reasoning, because the supplier can still perform. This distinction has been confirmed repeatedly in modern disputes, where tribunals have refused to treat a rise in cost, however painful, as an event that prevents performance (Rothstein, 2022). The reasoning is that a party who wanted protection against price swings could have bought it, through a fixed price, a hedge, an index clause, or insurance, and a court will not hand out that protection for free after the fact. 4.2 The Meaning of Hardship #hardship addresses a different situation. Here performance remains possible in a literal sense, but a change in circumstances has knocked the contract so far out of balance that demanding performance on the original terms would be deeply unfair to one side. The classic trigger is an event that either dramatically increases the cost of performance for one party or dramatically reduces the value of what the other party receives. The defining feature of hardship is imbalance rather than impossibility. A long-term supply contract priced in a currency that then collapses, a delivery obligation to a country that becomes commercially unreachable, or an energy contract signed before a fivefold jump in fuel prices are all classic hardship scenarios. The party can still perform, but the deal has become something neither side agreed to. Perales Viscasillas (2021) situates this idea within the older civil law doctrine of #rebus_sic_stantibus, which holds that a contract is binding only so long as the circumstances remain roughly as they were when it was made. The remedy for hardship is also different. Rather than suspension or release, the usual response is #renegotiation, and if that fails, either adaptation of the contract by a court or arbitrator or, as a last resort, termination. The idea is not to free the burdened party entirely but to restore a reasonable balance. This is a much more interventionist remedy than force majeure, and it is precisely because of this interventionism that many legal systems, especially in the common law world, have historically been reluctant to recognise hardship at all. A court that adapts a contract is, after all, imposing a bargain the parties never actually made, which sits uneasily with the idea that contracts are the parties' own creation. 4.3 The Core Distinction and Why It Matters The cleanest way to hold the two apart is to remember that force majeure is about whether you can perform, while hardship is about whether you should have to perform on the same terms. Force majeure looks at possibility. Hardship looks at fairness and balance. This distinction is not a mere academic nicety. It shapes what a party can ask for. A party facing a true impediment wants to be excused, and force majeure gives it that path. A party facing crushing but not impossible costs wants the deal adjusted, and only hardship offers that path. If a contract contains a force majeure clause but no hardship clause, a party drowning in extra costs may find that it has no remedy at all, because its problem is imbalance rather than impossibility. Qiu, Xu, and Yuan (2023) show that this gap is exactly where many pandemic disputes stalled, as parties tried to squeeze economic difficulty into a clause designed for physical impediment. The lesson for drafters is direct: a contract that anticipates only impossibility and ignores economic ruin is only half protected. 4.4 The Grey Zone Between Them In real life the two doctrines do not sit in tidy boxes. There is a grey zone where an event both obstructs performance and unbalances the deal, and where reasonable lawyers disagree about which label applies. A war, for example, might physically block a shipping route, which sounds like force majeure, while at the same time causing a huge spike in freight and insurance costs on the remaining routes, which sounds like hardship. Jevremovic (2022) points out that the same external event can, depending on how one frames its effect, be described either as an #impediment that prevents performance or as a change that merely makes performance more onerous. This overlap is one reason the doctrines are so often confused, and it is also why good drafting matters. A contract that clearly states which clause governs which kind of effect, and what happens in each case, spares the parties an expensive argument later. Where a contract is silent, the outcome may depend heavily on the governing law and on how a particular tribunal chooses to characterise the event. That uncertainty is itself a risk, and it is one that careful parties try to remove in advance rather than gamble on afterward. 5. Doctrinal and Historical Foundations 5.1 Pacta Sunt Servanda: The Default Rule At the base of all contract law sits a Latin maxim, #pacta_sunt_servanda, meaning agreements must be kept. This is the default, the gravity of the contractual universe. Without it, no one could rely on a promise, and trade would collapse. Every excuse doctrine is therefore an exception to this principle, and courts treat exceptions with suspicion. The starting assumption is always that a party who made a promise must keep it, even if keeping it has become inconvenient or expensive. Palmer (2022) describes how strict this principle once was. In the earliest common law, a promise was treated as nearly absolute. If you agreed to do something and then could not, that was your problem, because you could have protected yourself by writing an exception into the contract. An old and famous decision held a tenant still liable for rent even after being driven off the land by an invading army, on the reasoning that he could have bargained for protection and did not. This absolutism now seems harsh, but it explains why every later doctrine of excuse had to fight for its existence. The default did not disappear. It simply learned to bend a little in the most extreme cases. 5.2 Rebus Sic Stantibus and the Civil Law Counterweight Against pacta sunt servanda the civil law tradition developed a counterweight, the doctrine of #rebus_sic_stantibus, meaning things standing as they are. Under this idea, a contract carries an unspoken condition that it remains binding only so long as the surrounding circumstances stay roughly the same. If a fundamental change occurs, the binding force of the promise weakens, and relief may be available. This counterweight is the historical ancestor of the modern hardship doctrine. It reflects a different intuition from the common law starting point: that fairness sometimes requires adjusting a bargain when the world changes drastically. The tension between pacta sunt servanda and rebus sic stantibus runs through the entire subject. Too much respect for the first produces cruelty in extreme cases. Too much respect for the second produces uncertainty, because parties can never be sure their deal will hold. Every legal system, and every well-drafted contract, must find a balance between them, and the balance a system strikes tells you a great deal about how it thinks contracts should work. 5.3 Acts of God and the Common Law Path In the common law, the excuse doctrines grew slowly and cautiously. The phrase #acts_of_God originally described natural events with no human cause, and it gave early courts a narrow way to excuse performance made impossible by nature. Later the courts developed the doctrine of #frustration, which excuses a party when an unforeseen event destroys the whole basis of the contract or makes performance radically different from what was promised. A concert hall that burns down before a booked performance is the classic example, since the contract to hire it becomes pointless once the hall no longer exists. Alongside frustration, the common law recognised impossibility and, especially in the United States, #impracticability. Impracticability is slightly broader than impossibility, covering situations where performance is technically possible but only at an extreme and unreasonable cost caused by an unforeseen event. Even so, Palmer (2022) stresses that courts apply these doctrines narrowly, and that a simple rise in cost almost never qualifies. Frustration in particular is treated as a doctrine of last resort, not a convenient exit whenever a deal turns sour. The common law, in short, leans hard toward pacta sunt servanda and offers only a thin escape hatch. 5.4 Imprevision and the Civil Law Path Civil law systems took a more generous route. French law developed the doctrine of #imprevision, which allows relief and even judicial adaptation when an unforeseeable change makes performance excessively onerous. German law developed a related concept, often translated as the disturbance of the basis of the transaction, which allows a contract to be adjusted or, if adjustment is not possible, terminated. Several jurisdictions have recently written the hardship doctrine directly into their civil codes, giving courts explicit power to modify contracts when a change in circumstances makes performance excessively burdensome and the change was neither foreseeable nor assumed by the affected party. The contrast is striking. Where the common law asks mainly whether performance is possible, the civil law is more willing to ask whether the deal is still fair, and to step in and rebalance it if not. For a student, the practical takeaway is that the governing law of a contract can change the outcome dramatically. The same crisis may excuse a party under one system and leave it fully liable under another, purely because of the different attitudes toward hardship. This is not a small detail buried in the fine print. It can decide whether a company survives. 5.5 Excuse Doctrines as Risk Allocation Behind all these doctrines sits a single unifying idea: they are really about #risk_allocation. Every contract quietly assigns risks. When a buyer agrees to a fixed price, the buyer takes the risk that prices will fall and the seller takes the risk that they will rise. When a seller promises delivery by a certain date, the seller takes the risk that its own suppliers will be late. The excuse doctrines step in only when an event falls outside the risks the parties can be taken to have accepted, an event so far beyond the normal range that neither side truly priced it in. Seen this way, the narrowness of force majeure and the caution around hardship make sense. If courts handed out relief too freely, they would be re-allocating risks that the parties had already allocated, undoing the very bargain the parties struck. The doctrines exist for the residue, the genuinely extraordinary events, not for ordinary commercial misfortune. Keeping that purpose in mind helps a student predict outcomes, because it reframes the question from does this feel unfair to did the parties accept this risk. The second question is the one courts actually ask. 6. The International Instruments Because international contracts cross borders, parties often rely on harmonised instruments rather than on any single national law. Three of these matter most: the CISG, the UNIDROIT Principles, and the ICC model clauses. 6.1 The CISG and Article 79 The United Nations Convention on Contracts for the International Sale of Goods, usually called the #CISG, governs many cross-border sales of goods. Its central excuse provision is #Article_79, which releases a party from liability for damages if a failure to perform was due to an impediment beyond its control that it could not reasonably have been expected to foresee, avoid, or overcome. The provision also covers situations where the failure was caused by a third party engaged to perform, though the conditions there are stricter. Importantly, the exemption lasts only while the impediment lasts, and the affected party must give notice to the other side within a reasonable time. Article 79 sits closer to force majeure than to hardship. It speaks of an impediment, a word that suggests something blocking performance rather than merely making it expensive. The central and long-running debate, explored by Jevremovic (2022) and by Qiu, Xu, and Yuan (2023), is whether economic hardship can ever count as an impediment under Article 79. Some scholars argue that a severe enough imbalance should qualify. Others, and much of the case law, resist this, holding that a change in market conditions does not prevent performance and therefore does not excuse it. Advisory commentary on the Convention has noted that, in practice, tribunals have been extremely reluctant to exempt a party purely on grounds of economic hardship. Janssen and Wahnschaffe (2021) add that even where an impediment clearly exists, questions of timing and foreseeability often decide the case, because an event that was already unfolding when the contract was signed is not truly unforeseen. One further subtlety deserves attention. Article 79 excuses a party only from liability to pay damages. It does not automatically excuse the underlying obligation itself. This means a buyer might still, in theory, demand performance even where the seller is shielded from damages, a point that makes the provision narrower than many parties assume (Zeller, 2022). Because of these uncertainties, experienced drafters rarely rely on Article 79 alone. They write their own force majeure and hardship clauses to spell out exactly what happens, rather than leaving the outcome to a contested reading of the Convention. The CISG allows parties to do this, since its rules give way to what the parties themselves agree, and in practice most careful international contracts take advantage of that freedom. 6.2 The UNIDROIT Principles The UNIDROIT Principles of International Commercial Contracts, often used as a neutral reference or as the chosen law in international deals, are notable because they treat force majeure and hardship as two clearly separate ideas, each with its own rules. This separation is one of the great strengths of the instrument, and it models the clarity that national laws and private contracts often lack. The #UNIDROIT_Principles set out a force majeure provision that excuses non-performance when it is due to an impediment beyond the party's control that could not reasonably have been foreseen or overcome. This mirrors the classic three-part test and produces the classic remedies of suspension or release. The hardship provisions are separate and detailed. They define hardship as an event that fundamentally alters the equilibrium of the contract, either by raising the cost of performance or by reducing the value received, provided the event was unforeseeable, beyond the disadvantaged party's control, and not a risk that party had assumed. The remedy is a right to request renegotiation. If renegotiation fails, a court or tribunal may either terminate the contract or adapt it to restore its balance. This power of adaptation is exactly what the common law resists and what makes the UNIDROIT approach attractive to parties who want a fair, flexible outcome. For students, the UNIDROIT Principles are the clearest available illustration of the whole subject. They show, in a single instrument, how the two doctrines differ, what each requires, and what each delivers. A contract that borrows their structure is far less likely to fall into the common trap of confusing impossibility with imbalance. Even parties who do not choose the Principles as their governing law often use them as a drafting template, precisely because their treatment of these two doctrines is so clean. 6.3 The ICC Model Clauses The International Chamber of Commerce publishes model force majeure and hardship clauses that parties can drop into their contracts. These clauses are widely used because they are professionally drafted and internationally recognised. The ICC updated its model clauses in 2020, in the shadow of the pandemic, adding clearer references to events such as epidemics, currency restrictions, and trade sanctions, and offering different options for how a hardship situation should be resolved, including options that allow a tribunal to adapt the contract. The #ICC_model_clauses on force majeure typically list specific trigger events while also including a general catch-all test, so that an unlisted event can still qualify if it meets the core requirements of being beyond control, unforeseeable, and unavoidable. This hybrid design is deliberate. A pure list is safe but rigid, because anything not listed is arguably excluded. A pure general test is flexible but uncertain, because everything is open to argument. Combining the two gives parties both certainty for the obvious risks and a safety net for the unexpected ones. The ICC hardship clause gives parties a structured path: first attempt renegotiation, and if that fails, allow termination or, depending on the option chosen, judicial adaptation. Kiraz and Yildiz Ustun (2020) note that arbitral tribunals, when interpreting such clauses, focus heavily on the exact list of events and the precise wording, which is why the ICC's careful drafting is valuable. The broader lesson is that private clauses, whether from the ICC or drafted in-house, usually matter more than background law, because they replace vague doctrine with concrete instructions. A tribunal would much rather apply a clear clause than reconstruct what a distant statute or convention might mean for an unforeseen event. 6.4 Comparing the Instruments Placed side by side, the instruments reveal a spectrum. At one end sits the common law, cautious about excuse and deeply hostile to judicial adaptation. At the other end sit the civil law systems and the UNIDROIT Principles, which recognise hardship openly and allow a decision-maker to rebalance a contract. The CISG sits somewhere in between, generous about physical impediments through Article 79 but uncertain and largely silent about pure economic hardship. For anyone drafting an #international_contracts arrangement, this spectrum is the practical heart of the matter. The choice of governing law is not a formality. It quietly selects a whole philosophy about what should happen when the world changes. A party that wants the possibility of adaptation should lean toward a system or instrument that provides it, or should build that possibility into the contract expressly. A party that prefers certainty and a clean break should lean the other way. Either choice can be sensible. The mistake is to make the choice by accident, copying a governing-law clause from an old template without realising that it carries a hidden default about hardship along with it. 7. Stress-Testing the Doctrines Against Modern Crises The true test of any legal doctrine is how it performs when the world falls apart. The 2020s provided three natural experiments: a global pandemic, a major war with sweeping sanctions, and repeated supply-chain failures. Each stressed the doctrines in a different way. 7.1 The Pandemic The COVID-19 #pandemic was the largest single test of force majeure clauses in living memory. Governments closed borders, ordered businesses to shut, imposed quarantines, and disrupted transport. Companies around the world reached for their force majeure clauses to escape obligations they could no longer meet. The results were mixed, and the pattern is instructive. Where a government order directly and physically prevented performance, such as a mandatory factory closure or a total travel ban, force majeure claims often succeeded, because the event genuinely blocked performance (Natarajan, 2022). Where the pandemic merely made performance harder or more expensive, claims usually failed, because difficulty is not impossibility. A party that could still perform by paying more, finding another supplier, or absorbing a loss was generally expected to do so. Two further lessons emerged. First, wording was everything. Clauses that specifically mentioned epidemics, pandemics, or government action gave parties a clear path, while clauses that listed only wars and natural disasters left parties arguing about whether a virus counted (Rothstein, 2022). Second, #foreseeability shifted over time. A contract signed in early 2019 could plausibly claim the pandemic was unforeseeable. A contract signed in mid-2020, when the pandemic was already global news, could not, because by then the risk was obvious and the party was expected to have accounted for it (Janssen and Wahnschaffe, 2021). Bix (2022) cautions that even where relief seemed morally deserved, courts remained hesitant to rewrite bargains, preferring to hold parties to their words unless performance was truly impossible. The pandemic also exposed the gap between force majeure and hardship. Many businesses were not physically prevented from performing. They were simply ruined by the changed economics, facing collapsed demand or exploded costs. Their problem was imbalance, not impossibility, and force majeure could not help them. Only a hardship mechanism could, and many contracts did not contain one. This is perhaps the single clearest practical finding of the pandemic era, and it explains why the drafting of new contracts changed so quickly afterward, with hardship clauses appearing far more often than before. 7.2 War and Sanctions The war in Ukraine created a different kind of stress. Here the obstacles were partly physical, such as destroyed infrastructure, blocked ports, and active combat zones, and partly legal, in the form of #sanctions and export controls imposed by many governments. Both kinds of obstacle can trigger force majeure, but they raise distinct questions. Physical danger and destruction fit the classic force majeure model well. A supplier whose warehouse has been destroyed, or whose only shipping route runs through a war zone, has a strong claim that performance has been prevented. Legal obstacles are trickier. A sanction that makes it illegal to deal with a particular counterparty clearly prevents lawful performance, and most force majeure clauses that mention government action or changes in law will cover it (Zeller, 2022). But timing again becomes decisive. A party that entered a contract after sanctions were already in place, or after war was already underway, will struggle to argue that these events were unforeseeable. Sanctions also raise a harder question about duration. As a conflict drags on, events that once seemed extraordinary begin to look like the new normal. A party cannot keep invoking force majeure indefinitely for a situation that has become a predictable feature of doing business in a region. At some point the risk is treated as one the party has effectively accepted by continuing to trade. This is why careful drafters, when contracting in unstable regions, define the relevant events precisely and set clear limits, rather than relying on a general clause that may run out of force over time. It is also why some contracts distinguish between the outbreak of a new conflict, which may be excused, and the ongoing consequences of a known conflict, which may not. The war drove home the difference between force majeure and hardship once more. Some parties were not prevented from performing but were hit by enormous cost increases, particularly in energy, as markets reacted to the conflict and to sanctions. For them, the honest legal category was hardship, not impossibility, and their fate depended on whether their contracts contained a hardship mechanism or a governing law generous enough to provide one. Parties that had prepared only for physical impediment found themselves exposed to a purely economic shock. 7.3 Supply-Chain Failures The third crisis was less dramatic than war or pandemic but just as damaging: the repeated failure of global #supply_chain networks. A blocked canal, a port backlog, a shortage of a critical component, or a supplier's own collapse can ripple outward and leave a company unable to meet obligations it had every intention of keeping. Supply-chain cases are among the hardest to fit into force majeure, for two reasons. First, the disruption often does not make performance impossible but only more difficult or more expensive, which is the wrong side of the impossibility line. A buyer who can source the same goods elsewhere, even at a much higher price, is usually still expected to do so. Second, the failure is frequently caused by a third party, such as a subcontractor or upstream supplier, and both the CISG and many private clauses are strict about when a party can rely on someone else's failure as an excuse (Jevremovic, 2022). A party generally cannot hide behind a supplier's default unless that supplier would itself have been excused, and unless the party could not reasonably have found an alternative. The practical consequence is that supply-chain risk is best handled by drafting rather than by doctrine. Well-designed contracts now spell out what happens when a key input becomes unavailable, whether a party must seek alternative sources, how added costs are shared, and how long a suspension may last before either side can walk away. Contracts that leave these questions to a bare force majeure clause tend to produce disputes, because the clause was built for impossibility while the real problem is delay and expense. Here again, a hardship mechanism that allows adjustment and cost-sharing often serves parties better than a rigid excuse clause. The most resilient supply agreements treat disruption as a matter of when, not if, and build in the machinery to manage it before it happens. 7.4 Other Recurrent Shocks Although pandemics, wars, and supply-chain failures have dominated recent attention, the same analysis applies to other shocks that recur in modern commerce. Extreme weather and natural disasters, long the original home of the acts of God idea, remain frequent triggers, and they are becoming more so as climate patterns shift. Sudden changes in law, such as new export bans, tariffs, or import restrictions, can prevent lawful performance in the same way sanctions do. Cyber incidents that shut down a company's operations are a newer category, and whether they qualify depends heavily on whether the clause anticipates them and whether the affected party took reasonable precautions. The unifying point is that no list can ever be complete. New kinds of disruption keep appearing, which is why a good clause combines a specific list of known risks with a general standard for the unknown ones. Trying to predict every future crisis is hopeless. Building a clause that can respond sensibly to an unpredicted crisis is achievable, and it is the real goal of careful drafting. 8. Discussion: Drafting, Risk Allocation, and the Choice Between Excuse and Adaptation The crises of the 2020s point to one overriding conclusion: the outcome of a shock is decided far more by how a contract was written than by which doctrine a party later invokes. Doctrine is the background. The clause is the foreground. 8.1 The Primacy of the Clause In almost every dispute examined by recent scholars, the tribunal began not with abstract theory but with the exact words of the contract (Kiraz and Yildiz Ustun, 2020). A force majeure clause that listed epidemics protected its holder during the pandemic. One that did not left its holder arguing. A hardship clause that provided for renegotiation gave a struggling party a lifeline. Its absence left that party exposed. The words, in other words, did the real work. This is empowering for drafters. It means parties are not at the mercy of whatever the background law happens to say. They can build their own risk-allocation system, deciding in advance who bears which risks and what happens when each risk materialises. A contract is, at heart, a private allocation of future dangers. The excuse clauses are where that allocation becomes explicit, and treating them as an afterthought wastes the single best tool the parties have for protecting themselves. 8.2 Excuse or Adaptation The deepest choice a drafter faces is between excuse and adaptation. Force majeure offers excuse: when a qualifying event strikes, the obligation is suspended or released, and the parties largely go their separate ways. Hardship offers #adaptation: when a qualifying event strikes, the parties are pushed to renegotiate, and if they cannot agree, a neutral decision-maker may rebalance the deal so it can continue. Each approach suits different situations. For a one-off sale of goods, a clean excuse mechanism may be enough, because there is little relationship to preserve. For a long-term supply agreement, an energy contract, or a major construction project, adaptation is often wiser, because the parties want the relationship to survive the shock rather than shatter at the first sign of trouble. A well-drafted long-term contract therefore usually contains both a force majeure clause for genuine impediments and a hardship clause for severe imbalance, so that each kind of shock has its own tailored response. Guimaraes and colleagues (2021) confirm that the contracts that weathered the pandemic best were often those that had prepared for both possibilities. The choice between excuse and adaptation is, in the end, a choice about whether the relationship is worth saving. 8.3 Notice, Causation, and the Burden of Proof Even a perfect substantive clause can fail on procedure. Most force majeure and hardship provisions require the affected party to give #notice_requirement within a set time and in a set form, and a party that misses the deadline may lose the right to rely on the clause entirely, no matter how genuine the event. The notice obligation is not a technicality to be brushed aside. It exists so the other side can plan, mitigate, and make its own arrangements, and courts and tribunals take it seriously. The affected party also carries the burden of proof. It must show not only that a qualifying event occurred but that the event actually caused the failure to perform. This causation link is where many claims collapse. A party that would have breached anyway, for reasons unrelated to the crisis, cannot dress up its ordinary breach as force majeure. Rothstein (2022) notes that during the pandemic many parties invoked the crisis when their real problem was pre-existing weakness, and tribunals were quick to separate genuine impediment from convenient excuse. A student advising a client should therefore ask three questions in order: did a qualifying event occur, did it actually cause the non-performance, and was proper notice given. A weakness in any one of the three can sink the claim. 8.4 The Duty to Mitigate Neither doctrine rewards passivity. A party seeking relief is generally expected to have taken reasonable steps to avoid or reduce the effect of the event. This is the duty to mitigate, and it runs through the whole area. A supplier facing a blocked route is expected to look for another route. A buyer facing a failed supplier is expected to look for another supplier. Only after reasonable efforts have failed can the party plausibly say that performance was prevented or that the burden became excessive. Mitigation ties directly back to the impossibility line drawn earlier. What makes performance impossible is not just the event but the absence of any reasonable alternative. This is why cost matters in a subtle way: an alternative that exists only at a wildly disproportionate price may not count as a reasonable alternative, which is one of the few places where extreme expense can edge a case from mere hardship toward genuine impracticability. The boundary is narrow, and Palmer (2022) warns that courts patrol it carefully, but it exists, and it is worth understanding. 8.5 The Limits of Both Doctrines Neither doctrine is a cure-all, and students should resist the temptation to see them as magic escape routes. Force majeure is deliberately narrow, reserved for events that truly prevent performance, and courts guard that line carefully to protect pacta sunt servanda. Hardship is more flexible but also more uncertain, because it hands a decision-maker the delicate power to rewrite a bargain, a power that many systems grant reluctantly and use sparingly. Bix (2022) reminds us that the instinct to relieve suffering does not translate automatically into legal relief, and that both doctrines are shaped by a strong background commitment to holding parties to their promises. There is also a moral hazard concern that runs through the whole area. If excuse and adaptation are too easy to obtain, parties lose the incentive to plan, insure, and price risk properly, and contracts lose their reliability. If they are too hard to obtain, honest parties are crushed by events no one could have controlled. The doctrines, and the clauses that implement them, are constantly trying to strike this balance, and reasonable people disagree about exactly where the line should fall. That disagreement is not a flaw in the law. It reflects a real tension between certainty and fairness that no rule can fully dissolve. 9. Practical Guidance for Students and Drafters The following guidance distils the analysis into practical points that a student, a junior lawyer, or a business person can actually use. First, always separate the two questions. Before reaching for any clause, ask whether the problem is that performance has become impossible, which points to force majeure, or that it has become ruinously unbalanced, which points to hardship. Naming the problem correctly is half the battle. Second, write both mechanisms into long-term contracts. A serious cross-border agreement should contain a force majeure clause for genuine impediments and a separate hardship clause for severe imbalance. Relying on only one leaves a predictable gap that a future crisis will find. Third, be specific about trigger events. List the events that matter for your industry and region, including modern risks such as epidemics, sanctions, cyber incidents, extreme weather, and supply-chain breakdowns, while keeping a general catch-all so that unlisted events can still qualify if they meet the core test. The pandemic showed that a specific list is worth far more than a vague one. Fourth, address timing and foreseeability head on. State when notice must be given, how quickly, and in what form, because failure to give proper notice can sink an otherwise valid claim. Remember that an event known at signing is rarely unforeseeable, so a clause should make clear which risks the parties are consciously allocating. Fifth, decide in advance what the remedy is. Spell out whether a qualifying event leads to suspension, a duty to find alternatives, cost-sharing, renegotiation, judicial adaptation, or termination, and after how long. Do not leave the remedy to be argued about later, when the parties are already in conflict. Sixth, remember the third-party trap. If your ability to perform depends on suppliers or subcontractors, address explicitly what happens when they fail, because both the CISG and many clauses are strict about relying on someone else's default (Jevremovic, 2022). Seventh, choose the governing law with the doctrines in mind. Because common law and civil law systems treat hardship so differently, the choice of governing law can quietly decide the outcome of a future crisis. Make that choice deliberately, not by habit. Eighth, keep evidence and give notice promptly. When a crisis hits, document what happened, when, and how it affected performance, and send the required notice on time. A strong case can still be lost through poor record-keeping or a missed deadline. Ninth, do not treat these clauses as boilerplate. The single most expensive mistake of the last few years was signing standard clauses without reading them, then discovering during a crisis that they did not cover the crisis at hand. The few minutes spent reading and tailoring a clause can save a business. 10. Limitations and Future Research This article has deliberately taken a broad, comparative view, and that breadth comes at a cost. It does not resolve the many fine points of any single national law, and readers advising on a specific contract must consult the detailed rules of the governing system. The article also draws its examples chiefly from three recent crises, which, while highly relevant, may not capture every kind of disruption that future contracts will face. Several questions remain open and would reward further study. The first is whether pure economic hardship should ever qualify as an impediment under instruments like the CISG, a debate that recent scholarship has sharpened but not settled (Jevremovic, 2022; Qiu, Xu, and Yuan, 2023). The second is how far a decision-maker should be allowed to adapt a contract, and whether adaptation truly respects the parties' agreement or quietly replaces it. The third is empirical: as more arbitral awards and court decisions from the pandemic and the war become available, careful study of what actually persuaded tribunals would help drafters far more than abstract theory can. Future research that gathers and analyses these outcomes across systems would be especially valuable for students entering the field. 11. Conclusion Force majeure and hardship are two answers to two different questions, and the health of an international contract often depends on keeping them apart. #force_majeure asks whether an event beyond the parties' control has made performance impossible or has genuinely blocked it, and it responds with suspension or release. #hardship asks whether a change in circumstances has left the deal so unbalanced that fairness demands adjustment, and it responds with renegotiation or adaptation. The first protects against acts of God and their modern equivalents. The second protects against #economic_ruin when a bargain becomes something no one agreed to. The crises of the 2020s tested both doctrines hard. The pandemic showed that physical prevention is excused far more readily than economic difficulty, and that a clause is only as good as the events it names. The war in Ukraine and its sanctions showed that legal obstacles can trigger relief but that timing and duration steadily erode a party's ability to claim surprise. Supply-chain failures showed that delay and expense, however severe, rarely amount to impossibility, and that third-party defaults are treated with suspicion. Across all three, one lesson repeats: the words of the contract matter more than the doctrine behind them. For students and future practitioners, this is an encouraging conclusion. It means the tools of survival are largely in the parties' own hands. A contract that separates impossibility from imbalance, that names the risks that matter, that fixes remedies in advance, and that chooses its governing law with care is far better placed to survive whatever the future throws at it. The doctrines of force majeure and hardship are not obscure technicalities. They are, in a turbulent world, part of the basic machinery by which commerce keeps its promises when the ground shifts, and by which #international_contracts adapt rather than collapse. #force_majeure #hardship #acts_of_God #economic_ruin #pacta_sunt_servanda #rebus_sic_stantibus #CISG #Article_79 #UNIDROIT_Principles #ICC_model_clauses #frustration #impossibility #impracticability #imprevision #foreseeability #risk_allocation #renegotiation #adaptation #supply_chain_disruption #pandemic_contracts #war_and_sanctions #excused_performance #changed_circumstances #force_majeure_vs_hardship #navigating_force_majeure #international_commercial_contracts #cross_border_trade #global_supply_chains #contract_drafting #comparative_contract_law References Bix, B. H. (2022). COVID concerns: Some realism about equitable relief. Law and Contemporary Problems, 85(2), 37-49. Guimaraes, A. M. da Cunha, Souza Dantas Neto, M., Pugliesi, M., & Masseno, M. D. (2021). COVID-19 and the contractual performance crisis in international contracts governed by the Vienna Convention on Contracts for the International Sale and Purchase of Goods (CISG). Revista Juridica UNICURITIBA, 2(64), 158-183. Janssen, A., & Wahnschaffe, C. J. (2021). COVID-19 and international sale contracts: Unprecedented grounds for exemption or business as usual? Uniform Law Review, 25(4), 466-495. https://doi.org/10.1093/ulr/unaa026 Jevremovic, N. (2022). Article 79 CISG: Testing the effectiveness of the CISG in international trade through the lens of the COVID-19 outbreak. In P. Sooksripaisarnkit & D. Prasad (Eds.), Blurry boundaries of public and private international law: Towards convergence or divergent still? (pp. 127-155). Springer. https://doi.org/10.1007/978-981-16-8480-7_8 Kiraz, S. E., & Yildiz Ustun, E. (2020). COVID-19 and force majeure clauses: An examination of arbitral tribunals' awards. Uniform Law Review, 25(4), 437-465. Natarajan, P. (2022). May the force majeure be with you: The impact of COVID-19 on the force majeure clause in international commercial contracts. Santa Clara Journal of International Law, 21(1), 1-36. Palmer, V. V. (2022). Excused performances: Force majeure, impracticability, and frustration of contracts. The American Journal of Comparative Law, 70(Supplement 1), i70-i88. https://doi.org/10.1093/ajcl/avac017 Perales Viscasillas, M. del P. (2021). La rebus sic stantibus en el marco de la Convencion de Viena de 1980 sobre compraventa internacional de mercancias y en instrumentos de soft law. In S. Izaguirre Gomez, P. Perales Viscasillas, J. M. Elias de Tejada Casanova, & P. Lopez Martinez Acacio (Eds.), La rebus sic stantibus en tiempos de pandemia: Analisis general e impacto por sectores economicos (pp. 89-125). Tirant lo Blanch. Qiu, Z., Xu, Z., & Yuan, C. (2023). Spotlight on COVID-19: Analysis of the impediment under CISG Article 79 and suggestions for China. Journal of Education, Humanities and Social Sciences, 24, 405-411. https://doi.org/10.54097/6cbpay69 Rothstein, K. (2022). How parties can use COVID-19 to excuse performance of contracts. UC Davis Business Law Journal, 22(2), 297-336. Zeller, B. (2022). Covid-19 and Article 79: A revisit. In F. Benatti, S. Garcia Long, & F. Viglione (Eds.), The transnational sales contract: 40 years influence of the CISG on national jurisdictions (pp. 601-620). Wolters Kluwer and CEDAM.

The bribery of foreign public officials is a cross-border problem, yet the laws that punish it are still made and enforced by individual states. This article examines how two national statutes, the United States #Foreign_Corrupt_Practices_Act of 1977 and the #UK_Bribery_Act of 2010, reach conduct that happens far outside their own borders, and how that reach turns them into the working enforcement engines of the global fight against #corporate_bribery. The article treats these two statutes as essential companions to the #Merida_Convention, the popular name for the United Nations Convention against Corruption, which was opened for signature in Merida, Mexico, in December 2003. The Convention sets shared standards but leaves enforcement to national authorities, so the practical strength of the international regime depends heavily on a small number of active enforcers. Using a doctrinal and comparative method, the article explains the jurisdictional hooks of each statute, contrasts their treatment of #facilitation_payments and #corporate_liability, and describes the enforcement machinery behind them, including #deferred_prosecution_agreements and coordinated multi-country settlements. It then analyses recent shifts, including the 2025 pause and revival of #FCPA enforcement in the United States and the arrival of new failure to prevent offences in the United Kingdom, and considers what these mean for the durability of the system. The article closes with the main criticisms of #extraterritorial enforcement, from concerns about sovereignty and selective targeting to the risk that penalties rarely reach the countries where the harm was done, and draws out lessons for students and for compliance practice. Keywords: extraterritorial jurisdiction; foreign bribery; corporate criminal liability; anti-corruption law; deferred prosecution agreements; compliance. Introduction Corruption does not respect maps. A payment authorised in one country can be routed through a bank in a second, made to an official in a third, and recorded in the books of a parent company listed on a stock exchange in a fourth. This is the ordinary shape of modern #foreign_bribery, and it creates a hard question for the law. If the wrongdoing is spread across several states, which state gets to prosecute it, and on what basis? For most of the twentieth century the honest answer was that very few states prosecuted it at all. Paying a foreign official to win a contract was widely tolerated, sometimes even tax deductible, and treated as an unfortunate cost of doing business abroad. Two statutes changed that picture more than any others. The first was the #Foreign_Corrupt_Practices_Act, passed by the United States Congress in 1977 after investigations revealed that hundreds of American companies had made questionable payments overseas. The second was the #UK_Bribery_Act, passed in 2010 to modernise a patchwork of old British corruption offences and to bring the United Kingdom into line with its international commitments. Both laws share a striking feature. Neither is content to police only what happens on home soil. Each reaches conduct that occurs abroad, committed by companies and people who may have only a thread of connection to the enforcing country. This quality, the ability of a national law to govern conduct beyond national borders, is what lawyers call #extraterritorial reach. The central claim of this article is that these two statutes should be read as companions to the #Merida_Convention rather than as rivals or as purely domestic curiosities. The Convention, adopted by the United Nations General Assembly in October 2003 and entered into force in December 2005, is the only legally binding universal instrument against #corruption. It commits its many parties to criminalise bribery, to cooperate across borders, and to help recover stolen assets. What it does not have is a police force, a prosecutor, or a court of its own. It relies on states to act. In practice a handful of well resourced national enforcers, led by the United States and joined at times by the United Kingdom and a growing group of others, have supplied much of the muscle that the treaty itself lacks. Understanding the #FCPA and the #UK_Bribery_Act is therefore not a side topic for anyone studying international #anti_corruption. It is close to the heart of how the system actually works. The article proceeds as follows. Section 2 sets out the Merida Convention and explains why a binding treaty can still depend on national enforcers. Sections 3 and 4 examine the FCPA and the UK Bribery Act in turn, focusing on their structure and their jurisdictional reach. Section 5 compares the two directly. Section 6 describes the enforcement architecture, including negotiated resolutions and cross-border coordination. Section 7 covers developments between 2023 and 2026 that have reshaped the field. Section 8 returns to the companion theme, looking at mutual legal assistance, coordination, and the problem of double punishment. Section 9 gathers the main criticisms. Section 10 draws practical lessons, and Section 11 concludes. A note on method and scope is useful before going further. This article uses a doctrinal and comparative approach. It reads the two statutes, their guidance, and the leading cases against the text and structure of the Merida Convention, and it draws on recent scholarship and enforcement reporting to describe how the law works in practice rather than only on paper. It does not attempt an empirical study of its own. The focus is on the supply side of bribery, meaning the companies and individuals who pay, because that is where both statutes concentrate their force, though the demand side, meaning the officials who receive, is addressed where it matters. The aim is to give students a clear, accurate map of a fast moving field, and to make the case that the two national laws and the international treaty are best understood as parts of a single system. The Merida Convention as the global anchor The United Nations Convention against Corruption is usually shortened to UNCAC, but it is also known as the Merida Convention because it was first opened for signature at a conference in Merida, in the Mexican state of Yucatan, from 9 to 11 December 2003. It grew out of a recognition that earlier regional and sectoral instruments, useful as they were, could not match a problem that had gone fully global. The General Assembly adopted the text through Resolution 58/4 on 31 October 2003, and the treaty came into force in December 2005 once enough states had ratified it. It is now among the most widely joined treaties in the world, with membership covering the great majority of United Nations member states. The Convention is broad by design. It is built around five main areas: preventive measures, criminalisation and law enforcement, #international_cooperation, #asset_recovery, and technical assistance and information exchange. It asks states to establish anti-corruption bodies, to bring transparency to public procurement and public finances, and to criminalise a range of conduct that goes well beyond simple cash for contracts. That conduct includes the bribery of national and foreign officials, embezzlement, trading in influence, and the abuse of functions, along with corruption in the private sector. One of the treaty's genuine innovations was to make #asset_recovery a fundamental principle, an especially important point for developing countries that had watched public money disappear into foreign accounts. The structure of the Convention repays a closer look. Its chapters move from prevention, through criminalisation and law enforcement, to international cooperation and asset recovery, and finally to technical assistance and monitoring. The prevention chapter is often overlooked but is arguably the most ambitious, asking states to build honest public administration, to protect whistleblowers, to bring integrity to procurement, and to reach into the private sector by encouraging companies to keep proper accounts and adopt internal controls. The criminalisation chapter obliges parties to make bribery of national and foreign officials a crime and encourages them to reach related conduct such as concealment and laundering of the proceeds. To check on progress, the parties created a peer review process known as the Implementation Review Mechanism, in which countries examine one another's compliance. This mechanism can name gaps and prompt reform, but it cannot bring a prosecution or impose a penalty. It is a mirror, not a sword. That limitation is not a drafting failure; it reflects the political reality that sovereign states were willing to accept a common standard and mutual monitoring, but not a supranational body that could put their companies in the dock. The consequence, again, is that the treaty's real teeth are the national statutes of its more active members. For the purposes of this article, three features of the Convention matter most. First, it obliges parties to criminalise the bribery of foreign public officials, which lines up directly with what the FCPA and the UK Bribery Act do. Second, it presses states to take up broad bases of #jurisdiction and to cooperate through mutual legal assistance, extradition, and the sharing of evidence. Third, and this is the decisive point, it contains no supranational enforcement body. There is no international prosecutor for corruption and no standing court to try corporate bribery. The Conference of the States Parties and a peer review mechanism monitor how well countries are implementing their obligations, but monitoring is not the same as prosecuting. This design creates a gap between promise and practice. On paper, nearly every country has agreed to punish foreign bribery. In reality, enforcement has been concentrated in a small number of states with the will, the resources, and the legal tools to pursue complex international cases. Reviews of enforcement under the related OECD Anti-Bribery Convention have repeatedly found that only a minority of exporting countries actively investigate and sanction #foreign_bribery, while many do little or nothing. The Convention's ambition, in other words, has outrun the capacity of most of its members. This is exactly where the two national statutes come in. Because the FCPA and the UK Bribery Act reach conduct abroad and apply to a wide range of companies with even modest links to the United States or the United Kingdom, they end up doing much of the enforcement work that the treaty asks of all parties but that only a few actually perform. When a company is punished for a bribe paid in a third country, the practical effect is often to advance the Convention's goals through American or British law. Seen this way, the statutes are not competitors to the Merida framework. They are its enforcement companions, filling the space left by the absence of a global enforcer. The Foreign Corrupt Practices Act: origins, structure, and reach The #FCPA was born from scandal. In the mid-1970s, investigations connected to the Watergate affair uncovered that a large number of United States companies had maintained slush funds and made secret payments to foreign officials to win or keep business. Congress responded in 1977 with a statute that did two things at once. It made it a crime to bribe #foreign_officials, and it required companies to keep honest books and to maintain internal accounting controls. That second element, the #accounting_provisions, has turned out to be at least as powerful as the anti-bribery rule itself, because it does not require proof of an actual bribe. A company can be liable simply for hiding payments or failing to keep proper records. The FCPA has two main parts. The anti-bribery provisions prohibit offering, promising, or giving anything of value to a foreign official in order to obtain or retain business or to secure an improper advantage. The books and records and internal controls provisions apply to companies whose securities are registered in the United States, known as issuers, and require accurate record keeping and adequate controls. Enforcement is shared. The Department of Justice, the #Department_of_Justice, handles criminal cases and civil cases against non-issuers, while the Securities and Exchange Commission handles civil enforcement against issuers. This split matters, because the two agencies have different tools and different reach. The reach of the statute is what makes it a global instrument. Three categories of actor fall within it. First are issuers, meaning companies with securities listed or registered in the United States, including many foreign companies that raise capital there. Second are domestic concerns, meaning United States citizens, nationals, residents, and companies organised under United States law. Third, and most striking, are foreign persons and companies that take any act in furtherance of a corrupt payment while in the territory of the United States. Courts and enforcers have read these hooks generously. A wire transfer that passes through a United States bank, an email routed through a United States server, or a meeting held on United States soil can be enough to pull foreign conduct within reach. The result is that a bribe arranged entirely outside the United States, by a company with no obvious American identity, can still trigger #FCPA liability if the money or the communications touch the United States at some point. The accounting provisions widen the net further. Because they attach to issuers regardless of where the underlying conduct took place, a foreign company that lists in the United States must keep accurate global books. Payments disguised as consulting fees, commissions, or gifts can violate the record keeping rule even where the anti-bribery charge is hard to prove. In several of the largest cases, the accounting charges have carried much of the weight, and empirical research has linked the rise of extraterritorial enforcement, marked by greater international cooperation and heavier use of the accounting provisions, to real changes in corporate behaviour and even in foreign investment patterns in higher risk countries. A few definitions shape how far the statute stretches. The phrase anything of value is read broadly. It covers not just cash but gifts, travel, entertainment, job offers for a relative, charitable donations steered to please an official, and other benefits. The term foreign official is also wide. It includes officers and employees of a foreign government and of its agencies, but crucially it also reaches employees of state owned and state controlled enterprises. In many countries key industries such as energy, telecommunications, and construction run through such enterprises, so an ordinary commercial counterparty can turn out to be a foreign official for the purposes of the law. The requirement that the payment be made to obtain or retain business is likewise read to include securing any improper advantage, not only winning a specific contract. Each of these readings pushes the boundary of the statute outward and helps explain why cautious companies treat almost any dealing with the public sector abroad as a risk area. The penalties are heavy and layered. Companies face criminal fines, and the amounts can be increased under a general provision that ties penalties to the gain sought or the loss caused. On the civil side, issuers can be ordered to give up their ill gotten profits through disgorgement, together with interest. Individuals face fines and, for the anti-bribery offences, prison terms. On top of the numbers sit the softer costs already described: monitors, reporting duties, and reputational harm. The statute of limitations runs for several years and can be extended while evidence is gathered from abroad, which means that exposure lingers long after the conduct. A recent addition rounds out the American toolkit and closes an old gap. For most of its life the FCPA punished only the payer of a bribe, not the foreign official who demanded or took it. The Foreign Extortion Prevention Act, enacted in the United States in late 2023, now makes it a crime for a foreign official to demand or accept a bribe from a United States connected party. This turns attention to the demand side of #corruption, which the FCPA had left largely untouched, and it aligns American law more closely with the full scope of the Merida Convention, which addresses both giving and receiving. It is too early to judge how vigorously the new offence will be used, but its arrival shows that the American regime, even during a period of turbulence over the FCPA, has continued to grow in reach. One feature of the FCPA often surprises students. It permits, in narrow terms, so called #facilitation_payments, small sums paid to speed up routine, non-discretionary government actions such as processing a permit that the payer is already entitled to receive. This exception is narrow and easy to misjudge, and many companies now ban such payments outright because the line between a lawful facilitation payment and an unlawful bribe is thin and dangerous. Still, the exception exists in the American statute, and it is one of the clearest points of contrast with the British law, which allows no such thing. Who does the FCPA actually catch? For years, some of the very largest penalties were imposed on non-United States companies, a pattern that fuelled foreign complaints that the statute served American commercial interests. Whatever the motive behind particular cases, the practical effect has been to project a single country's anti-bribery rule across the globe, reaching German, French, Brazilian, and other companies whose only firm link to the United States was a stock listing or a dollar transaction. This is #extraterritoriality in its most visible form, and it is precisely what allows the statute to function as a companion to the Merida Convention, applying a shared standard to actors that their own governments were unwilling or unable to pursue. The UK Bribery Act 2010: structure and reach For a long time the United Kingdom relied on old and awkward corruption laws that were hard to apply to modern business. Pressure from international bodies, along with embarrassment over stalled investigations, pushed Parliament to act. The #UK_Bribery_Act, passed in 2010 and brought into force on 1 July 2011, replaced the old law with a clean and, in some respects, tougher framework. It is often described as one of the strictest anti-bribery statutes in the world, and the reasons are worth spelling out. The Act creates four offences. The first is the general offence of offering, promising, or giving a #bribe. The second is the general offence of requesting, agreeing to receive, or accepting a bribe, which means that being bribed is itself a crime, not only paying. The third is the specific offence of bribing a #foreign_officials, aimed at the same conduct that the FCPA targets. The fourth is the one that changed corporate life across the world: the offence, under section 7, of a commercial organisation failing to prevent bribery committed by an associated person for its benefit. Section 7 deserves careful attention. It is a form of #corporate_liability that does not require proving that senior managers knew about or approved the bribe. If someone who performs services for a company, an employee, an agent, a subsidiary, or another associated person, pays a bribe intending to benefit the company, the company itself commits an offence. The only defence is for the company to show that it had in place #adequate_procedures designed to prevent bribery. This flips the usual burden. Instead of prosecutors proving corporate guilt through the acts of a directing mind, the company must prove that it took reasonable preventive steps. The practical consequence has been enormous. Companies everywhere built compliance programmes, trained staff, screened third parties, and documented their controls, precisely so that they could rely on the #adequate_procedures defence if trouble came. Official guidance gives shape to what adequate procedures means. It sets out six principles that a company should weigh when building its programme: proportionate procedures matched to the risks the business actually faces; top level commitment from senior leaders who set the tone; risk assessment that is regular and documented; due diligence on the people and entities who act for the company; communication, including training, so that staff understand the rules; and monitoring and review so that the programme keeps pace with the business. These principles are deliberately flexible rather than a fixed checklist, because a small exporter and a global mining group face very different risks. The same six principle structure was later borrowed for the new failure to prevent fraud offence, which shows how influential the model has become. The definition of an associated person is what makes section 7 bite. An associated person is anyone who performs services for or on behalf of the company, which can include employees, agents, subsidiaries, distributors, consultants, and joint venture partners, judged by the substance of the relationship rather than its label. Because most foreign bribery is paid through intermediaries, this broad definition is exactly what allows the offence to reach the real world ways in which bribes are actually delivered. It also explains why third party management sits at the centre of any serious compliance effort. The penalties under the Act are severe. Individuals convicted of the general or foreign bribery offences can face up to ten years in prison and unlimited fines. Companies face unlimited fines, and a conviction can bring further consequences such as exclusion from public contracts. As with the American regime, the financial penalty is only part of the exposure; the disruption, the loss of management time, and the reputational damage can be just as costly. The Act's #extraterritorial reach is broad in its own way. The general offences and the foreign bribery offence apply to conduct anywhere in the world if committed by a person with a close connection to the United Kingdom, such as a British citizen, resident, or a company incorporated there. The section 7 failure to prevent offence goes further still. It applies to any commercial organisation that carries on a business, or part of a business, in the United Kingdom, no matter where the organisation is incorporated and no matter where the underlying bribery takes place. So a company based outside the United Kingdom, with only a portion of its operations there, can be prosecuted for failing to prevent a bribe paid in a completely different country by an agent who never set foot in Britain. Courts have interpreted the carrying on business threshold in a way that keeps the reach wide. Unlike the American statute, the UK Bribery Act allows no exception for #facilitation_payments. A small payment to speed up a routine act is, in principle, a bribe. Guidance recognises that businesses face real pressures in some countries and that prosecutors will use judgment, but the legal position is clear and unforgiving. This makes the British rule stricter on its face than the American one, and it is a common trap for companies that carry over an American style policy into a British legal environment. The Act is enforced primarily by the #Serious_Fraud_Office, a body that both investigates and prosecutes serious and complex economic crime. The Crown Prosecution Service and other agencies also have roles. The SFO's record has been mixed, with some large successes and some painful failures, a point examined further below, but its willingness to pursue foreign conduct with a British nexus has made the Act a serious presence in international enforcement rather than a dormant statute. Comparative analysis: two roads to the same destination Placed side by side, the FCPA and the UK Bribery Act reveal both a shared purpose and a set of instructive differences. Both aim to stop the #corporate_bribery of foreign officials, both claim reach beyond their borders, and both have reshaped global compliance. Yet the way they get there differs in ways that matter for companies and for scholars. Consider the jurisdictional hooks. The FCPA reaches issuers, domestic concerns, and foreign actors who take some act within United States territory. The UK Bribery Act reaches those with a close connection to the United Kingdom and, through section 7, any organisation that carries on business there. In broad terms the American reach leans on the presence of a United States link in the conduct or the corporate identity, often a listing or a dollar transaction, while the British reach leans on carrying on business in the United Kingdom. Both are wide, and in practice the two overlap constantly, which is why large multinational cases often draw in both sets of authorities at once. Consider corporate liability. The American approach has long rested on the doctrine that a company is responsible for the criminal acts of its employees and agents committed within the scope of their duties for the benefit of the company. This makes it comparatively easy to charge a corporation. The British approach traditionally required the prosecution to prove that a directing mind and will of the company, meaning a senior enough person to embody the corporate will, was personally culpable. That identification principle made corporate conviction hard, especially for large firms with diffuse decision making. Section 7 was the workaround for bribery, creating a strict corporate offence with a due diligence defence. This difference explains a lot about the two systems, including why negotiated resolutions took hold more quickly and more widely in the United States. Consider facilitation payments. The FCPA tolerates them in narrow circumstances. The UK Bribery Act does not. A single global company operating under both regimes must therefore write its policy to the stricter British standard, which is one reason many firms have abolished facilitation payments entirely. Consider the accounting dimension. The FCPA's #accounting_provisions give United States enforcers a powerful tool that does not depend on proving a bribe, only on proving false books or weak controls. The UK Bribery Act has no direct equivalent, though other British offences and the section 7 failure to prevent model achieve some of the same deterrent effect by pushing companies toward strong internal systems. The deeper point is that the two statutes are complementary rather than duplicative. Where one has a gap, the other often fills it. The American accounting provisions catch concealment; the British section 7 catches organisational failure to prevent. Together they raise the global cost of paying bribes and lower the payoff from lax controls. Empirical work supports this reading. Studies have found that the arrival of the UK Bribery Act had a measurable effect even on United States firms that were already subject to the FCPA, prompting exposed firms to pull back from higher risk markets. That finding is important because it shows that adding a second active enforcer does more than double the paperwork. It adds real deterrence at the margin, which is exactly what a treaty like the Merida Convention needs from its members but cannot itself deliver. The enforcement architecture: settlements, cooperation, and penalties Laws are only as strong as the machinery that enforces them, and the machinery behind these two statutes has become highly distinctive. Two features stand out: the heavy use of negotiated resolutions, and the growth of coordinated action across countries. In the United States, most corporate FCPA matters are resolved without a trial. Companies that discover wrongdoing often choose #self_reporting, cooperate with investigators, remediate their compliance failures, and then settle. The settlement may take the form of a #deferred_prosecution_agreements, a non-prosecution agreement, or a guilty plea, and it typically involves a large financial penalty, the disgorgement of profits, ongoing cooperation, and sometimes an external monitor. The logic is practical for both sides. Prosecutors secure a resolution, a penalty, and improved corporate behaviour without the cost and risk of trial. Companies avoid the catastrophic collateral consequences of a criminal conviction, which for some firms can mean debarment from public contracts or loss of licences. Critics argue that this bargain lets companies buy their way out of accountability, and that individuals too often escape, but the model has become the backbone of enforcement. The United Kingdom adopted its own version of the #deferred_prosecution_agreements in 2014, and it has become an important tool for the #Serious_Fraud_Office. A British DPA must be approved by a judge, who has to be satisfied that it is in the interests of justice and that the terms are fair, reasonable, and proportionate. This judicial supervision is a notable contrast with the more prosecutor driven American model. British DPAs have resolved several major matters, but the SFO's experience with individuals has been rocky. The identification principle made it hard to convict senior people, and there have been high profile acquittals even after the related company entered a DPA. This uneven record has driven much of the recent reform discussed in the next section. It is worth pausing on the corporate monitor, a device that turns a settlement into years of supervision. In some resolutions the company must accept an independent monitor, an outside expert who reviews its compliance programme, reports to the enforcer, and recommends improvements over a fixed term. For the company this is expensive and intrusive; for the enforcer it is a way to make sure that promised reforms actually happen rather than fading once the headlines pass. Monitors are not imposed in every case, and enforcers have at times signalled a preference for using them sparingly, but the possibility of one is a real part of the deterrent. The British system has used its own version of ongoing compliance obligations within the terms of a DPA. Since the British regime began in 2014 the Serious Fraud Office and, more recently, the Crown Prosecution Service have entered a modest number of agreements, far fewer than the American total, in part because the identification principle made it harder to build corporate cases in the first place. The December 2023 agreement with Entain, the first entered by the Crown Prosecution Service rather than the SFO, and resolving an alleged failure to prevent bribery in a legacy overseas business, showed that the tool is spreading to new agencies and new fact patterns. Two British cases illustrate the reach and the stakes. In the Airbus resolution of early 2020, bribery that occurred largely outside the United Kingdom was addressed as part of a coordinated settlement, showing how broadly the reach could be read. In late 2022, following a guilty plea, Glencore was ordered to pay a financial penalty of around 281 million pounds for bribery offences, including offences under section 7, as part of a wider global resolution that also involved authorities in the United States and Brazil. These cases show two things at once: the willingness of British authorities to pursue foreign conduct, and the increasingly coordinated nature of global settlements. Coordination is the second defining feature of the modern architecture. Large multinational cases now frequently involve several enforcers acting together, dividing penalties, aligning the facts admitted, and timing their announcements so that resolutions land at once. This coordination is partly practical, avoiding endless duplication, and partly principled, aiming to prevent a company from being punished many times over for the same conduct. It also depends on #international_cooperation, the sharing of evidence and mutual legal assistance that the Merida Convention and other instruments encourage. In this sense the daily grind of cross-border enforcement is where the treaty's cooperation provisions come to life. The penalties involved are large enough to change corporate behaviour. Fines and disgorgement in major matters run into the hundreds of millions and, in the biggest global resolutions, into the billions across all jurisdictions combined. For companies, the threat is not only the money. It is the monitor who sits inside the business for years, the reporting obligations, the reputational damage, and the risk to future contracts. For students of #enforcement, the important lesson is that the deterrent effect flows from the whole package, not from any single fine. Developments from 2023 to 2026: a system in motion Anyone studying this field must understand that it is not static. Between 2023 and 2026 both leading regimes shifted in ways that will shape enforcement for years, and the changes ran in opposite directions. In the United States, the most dramatic development came in early 2025. On 10 February 2025, the President signed Executive Order 14209, titled in terms of pausing #FCPA enforcement to further American economic and national security. The order directed the Attorney General, for a period of 180 days, to stop initiating new FCPA investigations and enforcement actions and to review and revise the guidelines that govern them. The stated rationale was that enforcement had become, in the administration's view, overexpansive and unpredictable, and had harmed American competitiveness. This was the first pause of its kind since the statute was enacted in 1977, and it sent a shock through the compliance world. Several points quickly became clear, and they are worth remembering because early headlines overstated the change. The statute itself was not repealed; only Congress can do that, so the FCPA remained the law. The pause applied to the Department of Justice, not to the Securities and Exchange Commission, which retained its civil authority over issuers. The statute of limitations, generally around five years, outlasts any single administration, so conduct during the pause could still be prosecuted later. And most conduct that violates the FCPA also violates other laws, such as wire fraud and money laundering, which lay outside the order. In short, the pause changed the posture of one agency for a period; it did not create a licence to bribe. The pause did not last. On 9 June 2025, a Deputy Attorney General issued new guidelines for FCPA investigations and enforcement. While the guidelines did not formally announce the end of the pause, they made clear that enforcement would move forward, though with new priorities and new internal controls. Those priorities included a focus on bribery connected to cartels and transnational criminal organisations, attention to conduct by foreign competitors that disadvantages American firms, and a requirement that new cases be authorised at a senior level. The practical message was that #FCPA enforcement was back, reshaped rather than abandoned. For companies, the safest reading was that abandoning compliance would be a serious mistake. The American wobble made the second development all the more important. While United States enforcement paused and then recalibrated, the United Kingdom moved in the opposite direction, toward more corporate accountability, not less. The vehicle was the Economic Crime and Corporate Transparency Act 2023, which received Royal Assent in October 2023 and came into force in stages. Two of its measures stand out. First, it reformed the identification principle for economic crimes, so that a company can be criminally liable where a senior manager, acting within the scope of their authority, commits a relevant offence. This directly addressed the long standing difficulty of convicting large companies. Second, it created a new corporate offence of #failure_to_prevent fraud, which came into force on 1 September 2025. The new failure to prevent fraud offence borrows the architecture of section 7 of the Bribery Act. A large organisation is liable if a person associated with it commits a specified fraud offence intending to benefit the organisation or its clients, unless the organisation can show it had reasonable fraud prevention procedures in place. It applies to large organisations, defined by thresholds for turnover, balance sheet total, and employee numbers, and it has #extraterritorial reach where there is a United Kingdom nexus, meaning the fraud takes place in the United Kingdom or the gain or loss occurs there. The offence extends the failure to prevent model well beyond bribery and tax evasion, and it signals a clear direction of travel in British policy. Alongside these statutory changes came institutional energy. The leadership of the #Serious_Fraud_Office signalled a more assertive stance, including an expectation that the new offences would drive more corporate resolutions. In March 2025, the United Kingdom, France, and Switzerland announced an International Anti-Corruption Prosecutorial Taskforce, a body designed to coordinate work on international bribery and corruption and open to other like minded agencies. The timing was notable. As one leading enforcer stepped back, others positioned themselves to fill the space, and the SFO indicated that it would pursue bribery allegations where United States authorities were not able to act. This is the companion dynamic in action: when one enforcer pulls back, the value of the others rises, and the overall system relies on a spread of active participants rather than a single dominant one. The lesson for students is that the global regime is not held up by any one country. It rests on a network of enforcers whose priorities can change with politics. The Merida Convention supplies the shared standard; the durability of enforcement depends on enough of its members choosing to act at any given time. Companions to Merida: cooperation, coordination, and the double punishment problem If the FCPA and the UK Bribery Act are the enforcement companions of the Merida Convention, the companionship shows up most clearly in three linked areas: mutual legal assistance, coordinated resolutions, and the effort to avoid punishing a company many times for a single course of conduct. The Merida Convention is not the only treaty in this story. The OECD Anti-Bribery Convention, which predates it, commits its parties, mostly major exporting economies, to criminalise the bribery of foreign public officials and subjects them to a rigorous peer review by a working group that publishes detailed country reports. The FCPA and the UK Bribery Act both sit within this overlapping treaty architecture, and the OECD review process has been one of the sharpest external pressures on member states to enforce their own laws. The reviews have also produced the uncomfortable finding, repeated over many years, that only a minority of parties enforce actively while many remain passive. This is the empirical backbone of the argument running through this article: shared standards are widespread, but real enforcement is concentrated. The two national statutes, backed by the OECD and Merida frameworks, are where the concentration mainly happens. Mutual legal assistance is the plumbing of cross-border enforcement. Investigators in one country need bank records, witness statements, and documents that sit in another. The Merida Convention, along with other treaties and bilateral arrangements, provides the legal basis for these requests. Without it, extraterritorial statutes would often be unenforceable in practice, because the evidence would be locked away in jurisdictions that had no obligation to help. The statutes supply the reach; the treaty framework supplies the means to gather proof. Neither works well without the other, which is another reason to see them as companions rather than as separate systems. Coordinated resolutions are the visible face of cooperation. In the largest cases, enforcers from several countries now align their investigations, agree on the facts a company will admit, split the financial penalties, and time their announcements to land together. This coordination reflects a shared recognition that a single global course of conduct should, so far as possible, produce a single coherent outcome rather than a series of uncoordinated blows. It also reflects the Convention's emphasis on cooperation, translated into everyday prosecutorial practice. The double punishment problem is the pressure point in all of this. If a bribe touches several countries, several of them may claim #jurisdiction. Without coordination, a company could face repeated prosecutions and repeated penalties for the same acts. Different legal systems handle the risk of double jeopardy differently, and the protection against being tried twice does not always cross borders cleanly. The practical answer has been the coordinated global settlement, in which enforcers credit penalties paid elsewhere and try to reach a proportionate total. This is not a perfect solution. Companies still worry about being pursued by a further authority after they thought a matter was closed, and the crediting of penalties across systems can be uncertain. But it is a workable answer, and it is one that grows directly out of the cooperative spirit of the Merida framework. There is a harder question lurking beneath the machinery. When a company bribes officials in a poorer country and is then fined hundreds of millions in the United States or the United Kingdom, where should that money go? Much of it stays with the enforcing state. The Merida Convention's #asset_recovery provisions were meant to steer stolen wealth back to the countries that lost it, but in practice the proceeds of foreign bribery penalties often do not reach the societies that suffered the underlying harm. This tension, between where enforcement happens and where the harm was felt, is one of the sharpest criticisms of the current system, and it is taken up next. Critiques and challenges Extraterritorial enforcement is powerful, but it is not free of serious objections. A fair account must set them out. The first objection concerns sovereignty. When one state prosecutes conduct that happened entirely within another state, committed by that other state's companies or nationals, it is, in a sense, exercising authority on foreign territory. Some governments have argued that broad #extraterritoriality intrudes on their right to regulate conduct within their own borders and to decide for themselves how to handle their own companies. The counterargument is that the enforcing state is acting on conduct with a genuine connection to itself, a listing, a transaction, or a business presence, and that the Merida Convention encourages exactly this kind of assertive jurisdiction. Still, the perception of overreach is real, and it carries diplomatic costs. The second objection is about selective enforcement and commercial advantage. Because the United States for many years imposed some of its largest penalties on foreign companies, critics abroad complained that the #FCPA had become a tool of economic competition dressed up as #anti_corruption. The framing of the 2025 executive order, which spoke openly about protecting American competitiveness and hinted at using enforcement against foreign competitors, gave those critics fresh material. If enforcement is steered by national economic interest rather than by an even handed commitment to fighting #corruption, then it starts to look less like the neutral policing of a global standard and more like the projection of national power. This concern cuts to the legitimacy of the whole enterprise. The third objection is the mismatch between where penalties land and where harm occurs. As noted above, the countries whose officials were bribed and whose public interest was damaged frequently see little of the money recovered. From their perspective, wealthy states are collecting large fines for wrongs suffered elsewhere. The Merida Convention's #asset_recovery provisions were designed to correct this, but implementation has lagged, and returning assets is legally and politically complicated. Until this improves, the system will remain open to the charge that it serves the enforcers more than the victims. The fourth objection concerns the reliance on negotiated resolutions. #Deferred_prosecution_agreements resolve cases efficiently, but they raise hard questions. Do they let companies treat penalties as a cost of doing business? Do they allow the individuals who actually paid or authorised bribes to escape while shareholders bear the fine? The British experience, where companies entered agreements but related individuals were sometimes acquitted, sharpened this worry and helped drive the reforms to the identification principle. There is a genuine debate about whether the model delivers real accountability or merely the appearance of it. The fifth challenge is inconsistency and instability. The 2025 pause showed how quickly the posture of a leading enforcer can change with a change in government. Companies that invested heavily in compliance found themselves guessing about enforcement priorities. Uneven enforcement across countries also creates an uneven playing field, the very thing these laws were meant to level. When only a few states enforce vigorously, companies from non enforcing states may gain an advantage, and companies from enforcing states may feel unfairly burdened. This is precisely why the spread of active enforcers, including the new taskforce, matters so much for the credibility of the regime. A final, subtler challenge concerns proof and complexity. Foreign bribery cases are among the hardest to investigate. Evidence sits abroad, witnesses are beyond easy reach, intermediaries obscure the money trail, and the conduct may be years old by the time it surfaces. These difficulties help explain why the #accounting_provisions and the failure to prevent model have become so central. Both reduce the need to prove the hardest facts, either by punishing concealment or by punishing organisational failure. That is a sensible response to a real problem, but it also shifts the law away from proving actual corruption toward proving weak systems, and reasonable people can disagree about whether that shift goes too far. Implications for students and for compliance practice For students, the study of these two statutes offers more than a set of rules to memorise. It offers a case study in how international law actually operates when there is no world government to enforce it. The Merida Convention shows that near universal agreement on a standard is possible. The FCPA and the UK Bribery Act show that turning that standard into real consequences depends on national institutions with the will and the tools to act extraterritorially. The gap between the two, between agreement and enforcement, is where much of the interesting work in this field lives. Several practical lessons follow for anyone who will work in or advise companies. The first is that reach is wide and connections are thin. A modest link, a dollar payment, a stock listing, a slice of business carried on in the United Kingdom, can be enough to bring foreign conduct within reach. Companies cannot assume that operating outside the United States or the United Kingdom keeps them safe. The second lesson is that #compliance is a legal defence, not just good manners. Under section 7 of the UK Bribery Act and under the new failure to prevent fraud offence, having reasonable or #adequate_procedures in place is the difference between liability and safety. A written policy that no one follows is worse than useless, because it shows awareness without action. The third lesson concerns third parties. In most foreign bribery cases the bribe is paid not by an employee directly but by an agent, a distributor, a consultant, or a joint venture partner. Both statutes reach conduct by associated persons, so companies must screen, monitor, and control the third parties who act for them. Due diligence on intermediaries is not a formality; it is often the single most important control. The fourth lesson is that #self_reporting and cooperation shape outcomes. The enforcement systems reward companies that find problems, report them, fix them, and cooperate, and they punish those that conceal. This changes the incentives inside a business, encouraging internal investigation rather than cover up. The fifth lesson is that #facilitation_payments are a trap. What is arguably permitted under the American statute is prohibited under the British one, so a global company should write to the stricter standard. The sixth lesson is that the ground shifts. The 2025 pause and revival, the arrival of new British offences, and the creation of a multinational prosecutorial taskforce all show that priorities move. Good compliance is built for the long run and does not chase the enforcement mood of a single year. Because the statute of limitations outlasts any administration, conduct that seems safe today can be prosecuted tomorrow. There is also a lesson about the ethics of the whole enterprise. #Corruption is not a victimless technicality. It distorts markets, raises the cost of public services, undermines trust in government, and falls hardest on people in the countries least able to bear it. The debates about sovereignty, selectivity, and where the penalties go are important, and students should take them seriously. But they are debates about how best to fight corruption, not about whether it should be fought. The core purpose of these laws, and of the Merida Convention behind them, is sound. Conclusion The bribery of foreign officials is a global problem that, for now, is policed mainly by national laws. The #Merida_Convention created a shared, binding standard and a framework for cooperation, but it left enforcement to states and gave itself no prosecutor and no court. Into that gap stepped the #Foreign_Corrupt_Practices_Act and the #UK_Bribery_Act. Through their #extraterritorial reach, their willingness to pursue conduct abroad, and the enforcement machinery built around them, these two statutes have done much of the practical work that the treaty asks of all its parties but that only a few actually perform. That is why they are best understood not as isolated national curiosities but as the essential enforcement companions of the Convention. The comparison between the two is instructive. The American statute leans on its #accounting_provisions and a comparatively easy route to #corporate_liability, and it tolerates narrow #facilitation_payments. The British statute leans on the failure to prevent model of section 7, refuses any facilitation exception, and now, through reforms in 2023, has strengthened both the identification principle and the failure to prevent approach across new areas. Where one has a gap, the other tends to have a tool, and together they raise the cost of #corporate_bribery worldwide. The recent history of the field is a reminder that the system is dynamic and, in some ways, fragile. The 2025 pause and revival of #FCPA enforcement showed how fast the posture of a leading enforcer can change, while the United Kingdom's new offences and the creation of a three country prosecutorial taskforce showed others stepping forward. The durability of the global regime does not rest on any single country. It rests on enough members of the Merida framework choosing to act, cooperating across borders, and refusing to treat #foreign_bribery as an acceptable cost of doing business. The serious criticisms of #extraterritorial enforcement, about sovereignty, selectivity, negotiated settlements, and the fate of recovered assets, deserve honest attention, and addressing them, especially the return of assets to the countries harmed, is the most important unfinished work. But the direction is clear. The fight against #corruption has gone global, and for the foreseeable future it will be carried on, case by case, through national statutes with worldwide reach acting as the working companions of a treaty that could not do the job alone. References Al-Hadi, A., Taylor, G., and Richardson, G. (2022). Are corruption and corporate tax avoidance in the United States related? Review of Accounting Studies, 27(1), 344 to 389. Bienenstock, S., and Kopp, P. (2023). The extensive reach of the FCPA beyond American borders: is a bad deal always better than a good trial? Public Choice, 196(3), 381 to 401. Cheung, Y. L., Rau, P. R., and Stouraitis, A. (2021). What determines the return to bribery? Evidence from corruption cases worldwide. Management Science, 67(10), 6235 to 6265. Christensen, H. B., Maffett, M. G., and Rauter, T. (2022). Policeman for the world: the impact of extraterritorial FCPA enforcement on foreign investment and internal controls. The Accounting Review, 97(5), 189 to 219. https://doi.org/10.2308/TAR-2020-0564 Escresa, L., and Picci, L. (2020). The determinants of cross-border corruption. Public Choice, 184(3), 351 to 378. Hock, B. (2020). Extraterritoriality and International Bribery: A Collective Action Perspective. London: Routledge. Kim, S., Rees, W., and Sila, V. (2020). Do anti-bribery laws reduce the cost of equity? Evidence from the UK Bribery Act 2010. Journal of Business Finance and Accounting, 47(3 to 4), 438 to 455. Organisation for Economic Co-operation and Development (2024). OECD Anti-Corruption and Integrity Outlook 2024. Paris: OECD Publishing. Sanseverino, A. (2024). The impact of foreign laws on U.S. firms: evidence from the U.K. Bribery Act. Review of Accounting Studies, 29(4), 3369 to 3412. https://doi.org/10.1007/s11142-023-09783-8 Transparency International (2022). Exporting Corruption 2022: Assessing Enforcement of the OECD Anti-Bribery Convention. Berlin: Transparency International. Topic hashtags #Foreign_Corrupt_Practices_Act #FCPA #UK_Bribery_Act #Merida_Convention #UNCAC #extraterritorial_reach #corporate_bribery #anti_corruption #foreign_bribery #deferred_prosecution_agreements #Serious_Fraud_Office #failure_to_prevent_bribery #adequate_procedures #asset_recovery #global_anti_corruption_enforcement

Foreign investment moves across borders on the strength of a quiet legal architecture that most people never notice. That architecture is built from #bilateral_investment_treaties, short agreements signed between two countries that promise to protect each other's investors. This article explains, in plain language, what these treaties are, where they came from, and how a single clause buried inside them gives a private company the power to drag a sovereign government before an international tribunal and demand billions of dollars. The clearest illustration is the Spanish energy firm Union Fenosa Gas, which used the 1994 Spain to Egypt investment treaty to win an award of more than two billion United States dollars against Egypt in 2018. Using that case as an anchor, the article maps the core protections found in almost every treaty, describes the private arbitration system known as #investor_state_dispute_settlement, and reviews the fierce debate over whether the system protects legitimate business or handcuffs the state's right to govern. It closes with the reform efforts now underway at the United Nations and in the European Union. The aim is to give students a solid, honest, and balanced foundation for understanding one of the most powerful and least understood parts of international economic law. Keywords: bilateral investment treaties, foreign direct investment, investor state dispute settlement, international arbitration, fair and equitable treatment, expropriation, ICSID, sovereignty, regulatory chill, treaty reform Introduction Every year, trillions of dollars in #foreign_direct_investment cross national borders. A German carmaker builds a plant in Mexico. A Chinese company buys a port in Greece. A Spanish utility signs a decades long deal to buy gas in Egypt. Behind each of these deals sits a worry that is as old as commerce itself. What happens if the host government changes the rules after the money is spent, seizes the asset, or simply refuses to keep its promises? For centuries the investor's options were poor. It could complain to its own government and hope for diplomatic pressure, or it could sue in the host country's own courts, where the same government that caused the harm often controlled the judges. #bilateral_investment_treaties, usually shortened to BITs, were designed to solve that problem. A BIT is a promise between two states that each will treat investors from the other fairly, will not take their property without paying for it, and will let them move their profits home. On its own, that promise would be ordinary. What makes BITs extraordinary is the enforcement mechanism attached to them. Most of these treaties allow the investor itself, not its home government, to sue the host state directly before a private international tribunal. This is the mechanism called #investor_state_dispute_settlement, or ISDS. It converts a diplomatic promise into a legal weapon that a company can fire on its own. The image of a spiderweb is useful here. Individually, each treaty is a thin thread linking two countries. Taken together, the thousands of treaties in existence form a dense and overlapping web that stretches across the entire planet. A single multinational company, by structuring its ownership through the right countries, can position itself so that many of these threads run through it. When a dispute arises, the company reaches for whichever thread gives it the strongest claim. The web is mostly invisible until something touches it, and then it vibrates all the way to a government treasury. This article is written for students who are meeting this topic for the first time. It assumes no prior knowledge of #international_law or #arbitration. Section two defines what a BIT actually is and what it contains. Section three traces the history of the regime and shows how it grew from a single treaty in 1959 into a global network of thousands. Section four opens up the anatomy of a typical treaty and explains the core protections that generate almost all disputes. Section five explains how the ISDS system works in practice. Section six examines the Union Fenosa case in detail, because it shows every part of the machine working together. Section seven surveys other landmark disputes that reveal both the reach and the controversy of the system. Section eight sets out the main criticisms and defenses. Section nine describes the reforms now being negotiated. The final sections draw the threads together and reflect on what the regime means for the balance between private capital and public power. What a Bilateral Investment Treaty Actually Is At its simplest, a BIT is a written agreement between two #sovereign_states in which each country agrees to give a defined set of legal protections to investors and investments coming from the other country. The typical treaty is short, often no more than fifteen pages, and follows a familiar template. It opens with a preamble stating that both countries want to encourage investment because investment promotes prosperity. It then defines its key terms, sets out the substantive protections, describes how disputes will be resolved, and states when the treaty enters into force and how long it lasts. Two definitions carry most of the weight. The first is the definition of an "investment." Older treaties define this very broadly, covering not only factories and shares but also loans, contracts, licenses, intellectual property, and concessions. The width of this definition matters enormously, because a claimant can only use the treaty if what it lost counts as a protected investment. The second key definition is that of an "investor," which usually covers both individuals who are nationals of a country and companies that are incorporated there. This second definition is the doorway through which a great deal of strategic behavior enters the system, because a company can often choose where to incorporate a subsidiary in order to gain the protection of a particular treaty. It is worth being precise about vocabulary, because the field is full of overlapping terms. The broad family of agreements that protect cross border investment is called #international_investment_agreements, or IIAs. Within that family, BITs are the largest group by number. The other main group is treaties with investment provisions, which are usually broad economic or free trade agreements that happen to include an investment chapter. The North American agreement that governed trade among the United States, Canada, and Mexico is a well known example of the second type. The distinction matters for lawyers, but for a student the important point is that a BIT and an investment chapter inside a trade deal usually do the same legal work. A BIT is a creature of #public_international_law, which means it is an agreement between states and is interpreted using the rules that govern treaties generally. Yet its practical beneficiaries are private companies, which is unusual. Most treaties create rights and duties only for the states that sign them. BITs are different because they give private actors a direct right to enforce state obligations. In legal terms this is sometimes described as granting investors "standing," meaning the right to bring a claim in their own name. Understanding this hybrid character, a public agreement with private enforcers, is the single most important step in understanding the whole system. Another feature that surprises newcomers is that BITs are almost always reciprocal on paper but lopsided in practice. The text treats both countries as equals, promising the same protections to investors from each side. In reality, capital tends to flow mainly in one direction, usually from a wealthier country to a developing one. A treaty between a large capital exporting economy and a small developing economy will, in practice, mostly be used by investors from the richer country suing the poorer one. This gap between formal equality and real world use lies at the heart of many criticisms discussed later. Finally, BITs are sticky. Many contain what is called a #survival_clause or sunset clause, which keeps the treaty's protections alive for investments already made even after the treaty is terminated, sometimes for ten, fifteen, or twenty additional years. A government that decides these agreements are a bad deal cannot simply walk away and be free the next morning. It remains exposed to claims from existing investors for a long time. This is one reason the web is so hard to dismantle. How the Web Was Woven: A Short History The modern story begins in 1959, when Germany and Pakistan signed the first recognizable bilateral investment treaty. In the years after the Second World War, European companies wanted to invest in newly independent countries but feared that their assets could be nationalized, as had happened in several waves of expropriation across the developing world. A treaty that promised compensation and protection was a way to reassure nervous investors and to attract the capital that developing countries said they wanted. The early treaties were, in this sense, a bargain. Poorer states offered legal protection in exchange for the promise of investment and the jobs and technology that were supposed to follow. For the first two decades the network grew slowly. The real turning point came in 1965, when the World Bank sponsored a multilateral agreement known as the #ICSID_Convention, formally the Convention on the Settlement of Investment Disputes between States and Nationals of Other States. This treaty created a dedicated institution in Washington, the International Centre for Settlement of Investment Disputes, to administer arbitration between investors and states. More than 160 countries have now joined it. ICSID did not by itself grant investors the right to sue. It provided the courtroom, so to speak, and waited for the individual BITs to hand investors the key that unlocked the door. The explosion came in the late 1980s and through the 1990s. The end of the Cold War, the spread of market oriented economic policy, and a global competition to attract capital produced a signing frenzy. Countries raced to conclude treaties, often using standard templates and giving little thought to the fine print, because signing a BIT was seen as a cheap and painless signal that a country was open for business. Many governments in this period signed dozens of treaties without a serious internal debate about the liabilities they were accepting. The result is that the great majority of treaties still in force today were drafted in the 1990s and 2000s, using broad and vaguely worded protections that later turned out to be far more powerful than their drafters imagined. The scale of the resulting network is striking. According to United Nations figures, by the end of 2024 there were roughly two thousand eight hundred #bilateral_investment_treaties in existence, alongside several hundred other treaties containing investment provisions, bringing the total universe of international investment agreements to more than three thousand three hundred. Of these, around two thousand six hundred were actually in force. This is a genuinely global system. Almost every country on earth is caught in at least a few threads of the web, and many are tied into dozens. For a long time the system was almost dormant. Treaties were signed but rarely used, because investors and their lawyers had not yet grasped how sharp a tool ISDS could be. That changed around the turn of the millennium. A handful of early arbitrations showed that the protections were real and that the awards could be large. Once the legal community understood this, the number of claims climbed steeply. The United Nations records fewer than six hundred known #ISDS cases at the end of 2013, but by the end of 2024 the total had risen to more than fourteen hundred, with fifty eight new cases filed in that single year. In roughly a decade the caseload more than doubled. To date, at least one hundred thirty five countries plus the European Union have been named as respondents in at least one claim. What began as a sleepy corner of international law had become one of its busiest and most consequential fields. The Anatomy of a Treaty: The Protections That Generate Disputes To understand why companies win such large awards, you have to look inside the treaty at the specific protections it grants. These clauses are short, sometimes a single sentence, but each has been interpreted by arbitration tribunals in ways that give it real bite. Six protections do almost all the work. The first and most important is #fair_and_equitable_treatment, universally abbreviated as FET. This clause requires the host state to treat foreign investors fairly. That sounds harmless, even obvious, but it has become the most powerful and most litigated standard in the entire system. Tribunals have read FET to protect an investor's "legitimate expectations," meaning the reasonable assumptions the investor made about how it would be treated when it decided to invest. If a government later changes a policy, revokes a permit, or fails to keep a promise in a way that upsets those expectations, it can be found to have breached FET even if it acted in good faith and broke no domestic law. FET is the clause invoked in the largest share of disputes, and it is the clause on which Union Fenosa won its case against Egypt. Because the words are so open ended, FET gives arbitrators wide discretion, and that discretion is exactly what makes it so valuable to claimants and so worrying to governments. The second protection is the rule against uncompensated #expropriation. Every treaty forbids the host state from taking an investor's property unless the taking serves a public purpose, is not discriminatory, follows due process, and is accompanied by prompt and adequate compensation. Direct expropriation, where the state formally seizes a factory or nationalizes an industry, is easy to recognize. The harder and more common category is #indirect_expropriation, sometimes called regulatory or creeping expropriation. This occurs when a government does not formally seize anything but adopts measures, such as new taxes, environmental rules, or license cancellations, whose effect is to destroy the value of the investment. The difficult question, argued in case after case, is where legitimate regulation ends and compensable expropriation begins. A rule that protects public health may look to a regulator like ordinary governance and to an investor like a taking of its business. The third protection is #most_favoured_nation treatment, or MFN. This clause says the host state must treat investors from the treaty partner no less favorably than it treats investors from any third country. Its original purpose was simple fairness, ensuring that no foreign investor is unfairly disadvantaged compared to another. In practice, clever lawyers have used MFN clauses to import better protections from other treaties. If country A's treaty with country B contains a weak protection, but country A's treaty with country C contains a stronger one, an investor from B may argue that the MFN clause entitles it to the stronger protection found in the C treaty. This technique, sometimes called treaty shopping through MFN, multiplies the effect of the web, because it lets threads borrow strength from one another. The fourth protection is #national_treatment, which requires the host state to treat foreign investors at least as well as its own domestic investors. This prevents a government from favoring local companies over foreign ones through discriminatory rules. The fifth is full protection and security, which obliges the state to provide physical and, according to some tribunals, legal security for the investment, for example by protecting it from mob violence or from a collapse of the legal order. The sixth protection is the #umbrella_clause. This clause, when present, requires the host state to observe any specific commitments it has made to the investor. Its effect is dramatic. It can lift an ordinary contract dispute, which would normally be handled by domestic courts, up into the treaty and turn a breach of contract into a breach of international law. An umbrella clause therefore blurs the line between a commercial disagreement and a treaty violation, and it broadens the range of conduct that can trigger a claim. Two further clauses complete the picture. The free transfer of funds provision guarantees that investors can move their capital and profits out of the country in a convertible currency, which protects them against capital controls. And, most importantly of all, the #dispute_settlement clause contains the state's standing offer to arbitrate. This is the clause that gives ISDS its force. By signing the treaty, the state makes an open offer to arbitrate any qualifying dispute. When an investor files a claim, it accepts that offer, and a binding arbitration agreement springs into existence even though the state and the investor never negotiated directly with each other. This is why the enforcement mechanism is often described as arbitration without privity. The consent is pre packaged inside the treaty, waiting to be triggered. The Engine Room: How Investor State Dispute Settlement Works The substantive protections would mean little without a way to enforce them, and enforcement is what sets this regime apart from almost every other area of international law. In most fields, a country that breaks an international rule faces, at worst, diplomatic pressure or complaints from other governments. Under a BIT, by contrast, the injured private company can go straight to a tribunal and ask for money, and the resulting award is enforceable against the state's assets around the world. The process usually begins when an investor believes a host state has breached one of the treaty protections. Many treaties require a cooling off period, often six months, during which the parties are supposed to try to settle. If that fails, the investor files a request for arbitration. Depending on the treaty, the case may be administered by ICSID in Washington, or under the rules of the United Nations Commission on International Trade Law, or by another arbitral institution. A tribunal is then formed, almost always of three arbitrators. In the typical model, the investor appoints one, the state appoints one, and the two sides agree on a president, or a neutral body appoints the president if they cannot agree. These arbitrators are not permanent judges. They are usually private practitioners, senior lawyers, and academics who are appointed case by case. This party appointed, case by case model is one of the most criticized features of the system, and we return to it later, but for now the key point is that a small, private, and temporary panel is deciding a dispute of enormous public consequence. The panel holds hearings, receives written pleadings and expert reports, and eventually issues a written decision called an award. Because the awards are reasoned and often published, they build up into a body of case law, even though there is no formal rule that one tribunal must follow another. The remedies available are almost always financial. A tribunal cannot usually order a government to reverse a law or reinstate a permit. What it can do is order the state to pay #compensation for the harm caused. The standard measure of damages in these cases is designed to put the investor in the position it would have been in had the breach never occurred. In practice this often means awarding the investor the lost future profits of the project, calculated using financial models that project decades of expected earnings and discount them to present value. This forward looking approach is why awards can be so large. The tribunal in the Union Fenosa case, for example, relied on the classic principle of full reparation drawn from a nineteen twenties decision known as the Chorzow Factory case, and awarded the investor more than two billion dollars representing lost profits. The sums involved are not trivial for the states on the receiving end. United Nations analysis reports that the average amount claimed by investors in these cases is around one point one billion dollars and the average amount actually awarded is around three hundred eighty five million dollars. In at least eight known cases, developing countries have been ordered to pay compensation of more than one billion dollars. For a wealthy state, a payment on that scale is painful. For a poor one, it can swallow a large share of the national budget or foreign currency reserves. One feature of the modern system deserves special mention because it has quietly changed the balance of power, and that is #third_party_funding. A specialized finance industry has grown up around investment arbitration. Outside investors, often hedge funds or dedicated litigation funders, will pay a claimant's legal costs in exchange for a share of any eventual award. This means that a company which could not otherwise afford a multi year, multimillion dollar case can still bring one, and it means that claims are increasingly assessed the way a financial product is assessed, in terms of expected return. Supporters argue that funding widens access to justice for investors with good claims but shallow pockets. Critics argue that it turns disputes into assets to be traded and encourages the filing of claims that might otherwise never have been brought. The reform negotiations discussed later treat third party funding as one of the systemic issues most in need of new rules, including disclosure requirements so that tribunals at least know who is standing behind a claim. Enforcement is the final piece, and it is what makes the awards impossible to ignore. Under the ICSID Convention, member states agree to recognize and enforce an ICSID award as if it were a final judgment of their own highest court. Awards issued outside ICSID are enforced through a separate 1958 treaty on the recognition of arbitral awards, which more than one hundred sixty countries have joined. Together these instruments mean that a winning investor can chase a losing state's commercial assets in courts around the world. This is why, in the Union Fenosa dispute, the company was able to pursue Egypt not only at ICSID but also in courts in the United States and England. A government cannot simply refuse to pay and expect the matter to disappear. Its assets abroad, from bank accounts to aircraft to real estate, may be exposed. Case Study: Union Fenosa Gas v. Egypt The dispute between Union Fenosa Gas and Egypt is worth studying closely because it puts every part of the machine on display, from the definition of an investment to the size of the final award and the long tail of enforcement litigation. It is also the example named in the title of this article, and it shows in concrete terms how a private energy company came to hold a two billion dollar claim over a national government. The background is a large energy project. Union Fenosa Gas was a Spanish company, a joint venture linked to the Spanish utility now known as Naturgy and the Italian energy group Eni. Through an eighty percent stake in an Egyptian company called SEGAS, it operated a liquefied natural gas plant at the port of Damietta on Egypt's Mediterranean coast. The plant was built to take Egyptian natural gas, cool it into liquid form, and export it by ship to international markets. Building the facility cost roughly one point three billion dollars, and its whole commercial logic depended on a long term supply of gas. To secure that supply, the project had a sales and purchase agreement under which an Egyptian state owned entity committed to deliver natural gas for at least twenty five years. Then the supply dried up. From around 2006 onward, Egypt began to divert gas away from export and toward its own domestic market, where demand and shortages were mounting. The problem deepened sharply during the political upheaval of 2011, when the country faced a severe economic and social crisis. Gas deliveries to the Damietta plant fell to levels so low that the facility was forced to shut down. From the investor's point of view, the whole basis of a very expensive project had been pulled out from under it by the actions of the state and its companies. In 2014, Union Fenosa Gas filed a request for arbitration at ICSID. It did not rely on the supply contract alone. It reached for a thread of the web, the 1994 agreement between Spain and Egypt on the reciprocal promotion and protection of investments, which is to say the Spain to Egypt #bilateral_investment_treaty. Because the company was Spanish and its investment was in Egypt, it fell squarely within the treaty's protection. This is the crucial move. A private commercial disagreement about gas deliveries was transformed, through the treaty, into a claim under #international_law that Egypt had failed in its duties toward a foreign investor. Egypt fought hard on several fronts. It argued that the tribunal had no jurisdiction, that there was no protected investment, and that the dispute was really just a contract matter that belonged in another forum. Most seriously, Egypt argued that the investment had been tainted by #corruption, claiming that the investor had secured its position through improper means. Corruption is a powerful defense in investment arbitration, because tribunals have held that an investment obtained through bribery does not deserve treaty protection. If Egypt had proved this, the case could have collapsed entirely. On 31 August 2018, the tribunal issued its award. It dismissed Egypt's objections and found that Egypt had violated its obligation to provide #fair_and_equitable_treatment to the Spanish investor. Applying the principle of full reparation, it concluded that the investor had lost more than two billion dollars in profits and ordered Egypt to pay, plus legal costs and interest. Reports at the time put the figure at over two billion dollars, with some accounts citing about two point two billion, making it one of the largest awards in ICSID's history. The three member tribunal was presided over by a widely respected British arbitrator, with one arbitrator appointed by each side. The arbitrator appointed by Egypt wrote a dissent, disagreeing in particular about the corruption question and pointing to what he saw as an unexplained gap between a local partner's role in the project and the money that partner received. That dissent is a reminder that these cases are often decided by fine judgment calls on which reasonable experts disagree. The story did not end with the award. Egypt sought to have the award annulled through ICSID's internal review process and fought a series of enforcement battles as Union Fenosa tried to collect. The company pursued recognition of the award in the United States and secured its registration before the courts of England. At one point a United States court noted, quite bluntly, the difficulty of taking so much money from a sovereign's treasury during a fiscal crisis that had been made worse by a global pandemic. The saga finally closed in March 2021, when the wider commercial dispute was settled as part of an agreement, led by Eni, to restart the idle Damietta plant, and the arbitration was formally discontinued. Several lessons stand out. First, the investor did not need its home government to act; it sued in its own name using a treaty signed decades earlier by politicians who could not have foreseen this dispute. Second, the size of the award flowed from the lost profits method, which valued a twenty five year gas contract as if it would have run its full course. Third, the corruption allegations show that these cases are not one sided formalities; states can and do mount serious defenses, and honest disagreement among arbitrators is common. Fourth, the years of enforcement litigation across three countries show how hard it is for a state to escape an award once it is issued. The Union Fenosa case is not an outlier or an abuse. It is the ordinary operation of a system working exactly as designed. The Web in Action: Other Landmark Disputes Union Fenosa is one strand among many. A short tour of other cases shows both the reach of the regime and the reasons it has become so controversial. The largest disputes tend to involve natural resources and energy, where projects are huge and the stakes enormous. In Tethyan Copper Company v. Pakistan, an Australian investor sought to develop a vast copper and gold deposit called Reko Diq in the province of Balochistan. When the provincial authorities refused to grant the mining lease, the investor brought a claim under the Australia to Pakistan treaty. In 2019 an ICSID tribunal found that Pakistan had breached #fair_and_equitable_treatment and had unlawfully expropriated the investment, and it ordered Pakistan to pay damages of roughly five point eight billion dollars. For a country facing economic strain, an award of that size was staggering, equal to a large slice of its foreign reserves. The dispute was eventually resolved through a settlement in 2022 that revived the project under new arrangements, but the case remains a warning about how a single permit decision can create a multibillion dollar liability. The single largest award ever recorded arose from the #Energy_Charter_Treaty, a multilateral agreement focused on the energy sector. In a set of claims brought by the former majority shareholders of the Russian oil company Yukos, an arbitral tribunal in 2014 ordered Russia to pay around fifty billion dollars, concluding that the company had been dismantled through tax and enforcement measures amounting to an unlawful taking. The award has been the subject of years of court battles over its validity and enforcement, but the sheer number, fifty billion dollars against a major state, shows the scale that this system can reach. Other cases became famous not for their size but for what they revealed about the tension between investment protection and public policy. When Australia introduced plain packaging rules for cigarettes to protect public health, the tobacco company Philip Morris restructured part of its business through Hong Kong and brought a claim under an Australian investment treaty. In 2015 the tribunal threw the claim out, finding that the corporate restructuring had been done mainly to gain access to the treaty and therefore amounted to an abuse of the process. A parallel claim against Uruguay over similar tobacco rules was decided in 2016, and there the tribunal ruled in favor of Uruguay, upholding the state's right to regulate for public health. These tobacco cases are often cited by both sides. Defenders say they show the system can reject weak or abusive claims. Critics say the mere fact that a government had to spend years and millions of dollars defending a public health law is itself the problem. Environmental and energy transition measures have produced a growing wave of disputes. In the early NAFTA era, the case of Metalclad v. Mexico established that local measures blocking a hazardous waste facility could amount to indirect expropriation, an early signal that environmental regulation could trigger liability. More recently, Germany's decision to phase out nuclear power after the Fukushima disaster led to a long claim by the Swedish energy company Vattenfall, which was ultimately resolved through a settlement in 2021. Ecuador was ordered to pay the oil company Occidental well over a billion dollars in a dispute arising from the termination of an oil contract. Across these cases a pattern appears. When a government changes course on energy, resources, or the environment, foreign investors who bet on the old policy may seek to be paid for the change. United Nations figures show that around a third of all known disputes involve the energy and extractive sectors, with more than two hundred cases tied to fossil fuels and over a hundred connected to renewable energy. These examples also reveal the strategic use of the web. The Philip Morris restructuring through Hong Kong, and the general practice of routing investments through countries with favorable treaties, is known as #treaty_shopping. A company that wants the protection of a particular treaty can often obtain it simply by placing a holding company in the right jurisdiction before a dispute arises. This is legal in many circumstances, and it means that the map of who can sue whom is not fixed by nationality but can be engineered. The spiderweb, in other words, is not only inherited; it can be redesigned by those who know how to use it. The Great Debate: Criticisms and Defenses Few areas of international law are as contested as the investment treaty regime. The arguments cut deep because they touch the boundary between the rights of capital and the power of democratic states. A fair account must present both sides. The case in favor is straightforward. Supporters argue that BITs #depoliticize investment disputes. Before the regime existed, an investor whose property was seized had to persuade its home government to intervene, which turned commercial harm into a matter of foreign policy and sometimes into gunboat diplomacy. A neutral tribunal is a great improvement over that. Supporters also argue that credible protection attracts capital. A developing country that can promise investors a fair and enforceable set of rules should, in theory, be able to attract more investment on better terms than a country whose courts are weak or captured. And they argue that the rule of law itself benefits, because the discipline of possible claims encourages governments to treat all investors, foreign and domestic, more predictably and more fairly. The criticisms are equally serious and have grown louder over time. The first is the charge of #regulatory_chill. This is the concern that governments, fearing expensive claims, will hesitate to pass laws that protect health, the environment, or workers. The chilling effect is hard to measure precisely, because it involves laws that were never passed, but a substantial body of scholarship argues that the fear of arbitration does influence policy choices, especially in developing countries with limited legal budgets. When defending a single case can cost millions of dollars in legal fees regardless of whether the state wins, even a weak claim imposes a real cost, and the shadow of that cost can fall across the whole process of governing. The second criticism concerns the structure of the tribunals themselves. Because arbitrators are appointed case by case, and because the same individuals often serve as arbitrator in one case and as counsel in another, critics argue that the system lacks the independence and impartiality expected of real courts. There is no security of tenure, and some worry that arbitrators have a quiet incentive to keep the system, and the flow of cases, alive. Related to this is the problem of #inconsistency. Different tribunals interpreting similar clauses have reached opposite conclusions, and because there is no true appeal and no binding precedent, the law can be unpredictable. Two investors in nearly identical situations can receive opposite outcomes. The third criticism is about asymmetry and sovereignty. Under most treaties, only the investor can sue. The state cannot bring an affirmative claim against a company that pollutes or breaks its promises; it can only defend. Critics see this as a one way ratchet that grants rights to capital without matching duties. More fundamentally, they argue that the system allows a private panel to sit in judgment over the sovereign acts of a democratically elected government, and to attach a price tag to public policy. When a tribunal awards billions against a state for a decision made in the public interest, the question of who really governs becomes sharp. The fourth criticism is distributional. The system's costs fall hardest on the countries least able to bear them. United Nations data show that the majority of claims are brought against developing countries, very often by investors from wealthy ones. The formal reciprocity of the treaties does not translate into real world balance, because capital mostly flows one way. When a poor country is ordered to pay a billion dollar award, the money that leaves its treasury is money not spent on schools, hospitals, or infrastructure. The regime that was sold as a tool of development can, in these cases, work against it. There is also a growing tension with climate policy. As governments try to move away from fossil fuels, investors who own coal plants, oil concessions, and gas projects hold treaty rights that may entitle them to compensation when those assets are shut down. Some scholars warn that the threat of large claims could slow the energy transition at exactly the moment when speed matters most. Others respond that well drafted treaties can carve out space for climate action, and that the problem lies in old treaties rather than in the concept itself. This debate is now one of the most active in the field. It helps to look at the actual outcomes rather than the loudest headlines. Across the full body of decided cases, the results are more mixed than either camp usually admits. Investors do not win every time. A large share of claims are dismissed on jurisdiction or lost on the merits, and many are settled on confidential terms before any award. But when investors do win, the amounts can be very large, and even a successful defense leaves the state with a heavy legal bill and years of distraction. So the raw win rate understates the pressure the system places on governments, because the cost and uncertainty of being sued at all is a burden regardless of the final verdict. This is why measuring the fairness of the regime is so difficult. Counting only who wins misses the chilling effect of who might be sued, and counting only the largest awards misses the many cases that fizzle out quietly. An honest assessment has to hold the whole distribution of outcomes in view, from the record breaking awards down to the claims that fail. None of these criticisms proves that the system is worthless, and none of the defenses proves that it is fine as it is. The honest position, and the one most scholars now hold, is that the regime delivers real benefits and real harms at the same time, and that its particular design, written mostly in the 1990s, is poorly suited to the pressures of the present. That recognition is what has driven the reform movement. Repairing the Web: The Reform Movement The debate has moved beyond talk. Over the past decade a serious, if slow, effort has been underway to reform the system, and it is happening on several fronts at once. The most ambitious effort is taking place at the United Nations. Since 2017, a body known as Working Group Three of the United Nations Commission on International Trade Law has been meeting to consider possible reform of #investor_state_dispute_settlement. Its members, drawn from states around the world, have identified a shared list of concerns, including the inconsistency of decisions, doubts about arbitrator independence, the high cost and long duration of cases, and the difficulty developing countries face in handling claims. The Working Group has produced concrete results. It agreed on a Code of Conduct for arbitrators and judges in these disputes, intended to tighten the rules on conflicts of interest. It approved the statute for an Advisory Centre designed to help poorer states, which often cannot afford top tier legal representation, to defend themselves. And it has been drafting procedural reforms and debating a possible multilateral instrument that would allow many countries to update their old treaties at once. The hardest question at the United Nations is whether to replace the current model of party appointed, case by case tribunals with a permanent standing body. The European Union has pushed hardest for this idea, proposing a #multilateral_investment_court with tenured judges and a genuine appeal mechanism, which would address the twin problems of inconsistency and independence. Other states prefer to keep arbitration but improve it around the edges. As of the mid 2020s the negotiations continue, with agreement on some incremental fixes but no consensus yet on the deepest structural change. Reform by committee across dozens of sovereign states is a patient business. A second front is the wave of terminations and renegotiations. Some countries decided that the old treaties were simply not worth their risks and began to exit. A number of states have terminated older BITs, and the network is slowly shrinking at the edges. Within Europe, a dramatic step followed a 2018 ruling by the European Union's highest court, in the Achmea case, which held that arbitration clauses in treaties between European Union member states were incompatible with European Union law. In 2020, most member states signed an agreement to terminate their treaties with one another, unwinding a large cluster of the web from the inside. In parallel, the European Union and its members moved to withdraw from the Energy Charter Treaty after efforts to modernize it stalled, citing its incompatibility with climate goals. A third front is quieter but may matter most in the long run, and that is the drafting of new treaties. A new generation of #international_investment_agreements is being written with the lessons of the past in mind. Instead of the open ended fair and equitable treatment clause that gave tribunals so much discretion, many recent treaties replace it with a closed and specific list of what the state must and must not do. United Nations analysis finds that roughly half of the treaties signed with protection standards in the past five years take this narrower approach. New treaties increasingly include explicit language preserving the state's #right_to_regulate in the public interest, carve outs for health and environmental measures, and tighter definitions of investment designed to block treaty shopping. Some drop investor state arbitration entirely and send disputes to state to state channels or domestic courts. The trouble is that these improvements affect only a small slice of the total exposure. The same United Nations data show that only about sixteen percent of the world's stock of foreign investment is covered by these new, reformed agreements. The overwhelming majority of investment is still governed by the old treaties from the 1990s and 2000s, with their broad standards and easy access to arbitration. Because of survival clauses, even terminated treaties keep protecting existing investments for years. Reforming the future is easier than disarming the past, and the web woven decades ago will keep vibrating for a long time to come. Discussion: What This Means for Students and Citizens Stepping back, what should a student take from all of this? The first lesson is that law shapes power in ways that are often invisible. A short treaty signed by two governments in 1994, using standard language and attracting no public attention at the time, gave a Spanish company the ability, twenty years later, to win two billion dollars from Egypt. The people affected by that outcome, the Egyptian public whose government had to account for the liability, never voted on the treaty and, in most cases, never knew it existed. This gap between the quiet making of legal rules and their loud consequences is a theme that runs through much of international economic law. The second lesson is that neutral sounding words carry enormous weight. Phrases like #fair_and_equitable_treatment or protection against expropriation seem almost too obvious to argue with. Who could be against fairness? Yet it is precisely in the interpretation of these mild phrases that billions of dollars change hands and the balance between private rights and public power is set. Students of law and policy should learn to read such clauses with suspicion, asking not what they seem to say but what they have been made to mean in practice. The third lesson is that the debate is not a simple contest between good and evil. It is tempting to cast investors as villains and states as victims, or the reverse. The reality is that the system was built for a reasonable purpose, to make cross border investment safer and less political, and that it has delivered some of what it promised while also producing outcomes that many people find hard to accept. Holding both of these truths at once, seeing benefit and harm together, is the mark of a mature analysis. The reform movement exists precisely because thoughtful people concluded that the answer is neither to worship the system nor to burn it down, but to redesign it. The fourth lesson is about #sovereignty in a connected world. States remain the primary actors in international affairs, but they operate inside a web of commitments that limits what they can do. Every treaty a government signs trades away a little freedom of action in exchange for some expected benefit. The investment regime is a vivid example of that trade, because the loss of freedom is so concrete and the price of exceeding it is so precisely measured in money. Understanding this helps explain why governments are now so cautious about what they sign, and why unwinding old commitments is so difficult. For a student who wants to go further, the practical skills are clear. Learn to read a treaty text closely. Trace how a general clause has been applied in specific awards. Follow the money in a damages calculation to see how future profits become present liabilities. And watch the reform negotiations, because the rules being written now will govern the investments of the coming decades. The field rewards those who can combine legal precision with an honest view of the political stakes. Conclusion Bilateral investment treaties are the hidden scaffolding of the global economy. Individually modest, collectively vast, they form a spiderweb of more than two thousand eight hundred agreements that ties almost every country into a shared set of promises about how foreign investors will be treated. The threads are usually invisible, but when a dispute touches one, the whole structure responds, and the vibration can reach all the way into a national treasury. The genius, and the danger, of the system lies in one design choice, the decision to let private companies enforce these promises directly through binding international arbitration. That choice converts a diplomatic courtesy into a legal weapon. It is why a Spanish gas company could take a sovereign state to a tribunal in Washington and walk away with an award of more than two billion dollars, and why similar claims have produced awards running into the tens of billions elsewhere. The Union Fenosa case is not an aberration but a clear window into the normal working of the machine, from the broad definition of investment, through the powerful fair and equitable treatment standard, to the lost profits damages and the years of global enforcement. Whether this system is a triumph or a trap depends on where one stands and which case one examines. It has depoliticized disputes and offered investors real protection. It has also chilled regulation, strained public budgets, and placed private panels in judgment over public policy, most heavily in the countries least able to bear the cost. The current wave of reform, from the United Nations negotiations to the new generation of narrower treaties, is an admission that the old design no longer fits the world it governs. But because the web was woven decades ago, and because its threads survive even after treaties are cut, the reform will take a generation to work through. For students entering this field, that slow unwinding is not a reason to look away. It is an invitation to understand, and perhaps to help redraw, one of the most consequential and least visible legal structures of our time. Hashtags #Bilateral_Investment_Treaties #BITs #Investor_State_Dispute_Settlement #ISDS #Foreign_Direct_Investment #International_Investment_Law #Union_Fenosa_v_Egypt #ICSID #Fair_And_Equitable_Treatment #Expropriation #Sovereignty_And_Regulation #Regulatory_Chill #International_Arbitration #Treaty_Reform #Global_Investment_Governance References Choiniere, H., and Maksimov, V. (2022). Investor state dispute settlement and sustainable development: negative externalities and a need for reform. AIB Insights, 22(1). https://doi.org/10.46697/001c.32976 Di Salvatore, L. (2021). Investor state disputes in the fossil fuel industry. Winnipeg: International Institute for Sustainable Development. Dolzer, R., Kriebaum, U., and Schreuer, C. (2022). Principles of International Investment Law (3rd ed.). Oxford: Oxford University Press. Hagemann, T. (2022). The north south divide of regulatory chill: a comparative analysis of the impact of investor state dispute settlement on policy makers in developed and developing countries. New York International Law Review, 35(1). Horn, H., and Tangeras, T. (2021). Economics of international investment agreements. Journal of International Economics, 131, 103433. Paine, J., and Sheargold, E. (2023). A climate change carve out for investment treaties. Journal of International Economic Law, 26(2). Reinisch, A., and Mansour Fallah, S. (2022). Post termination responsibility of states? The impact of amendment, modification, suspension and termination of investment treaties on vested rights of investors. ICSID Review, 37(1-2), 101-120. https://doi.org/10.1093/icsidreview/siab023 Salacuse, J. W. (2021). The Law of Investment Treaties (3rd ed.). Oxford: Oxford University Press. Schreuer, C., Malintoppi, L., Reinisch, A., and Sinclair, A. (2022). The ICSID Convention: A Commentary (3rd ed.). Cambridge: Cambridge University Press. United Nations Conference on Trade and Development. (2024). Facts and Figures on Investor State Dispute Settlement Cases, IIA Issues Note No. 3, 2024. Geneva: United Nations. United Nations Conference on Trade and Development. (2024). World Investment Report 2024: Investment Facilitation and Digital Government. Geneva: United Nations. United Nations Conference on Trade and Development. (2025). International Investment Agreements: Trends and Recent Developments, IIA Issues Note No. 3, December 2025. Geneva: United Nations.

Few ideas in #international_arbitration cause as much confusion, and as much avoidable litigation, as the difference between the seat and the venue. To a newcomer the two words sound like synonyms. In law they are not. The #legal_seat is the juridical home of an arbitration. It fixes the #procedural_law that governs the reference, decides which national courts hold #supervisory_jurisdiction, gives the resulting award a nationality, and shapes how and where that award can be set aside or enforced. The #physical_venue is only the geography of the process, the rooms and cities where hearings, meetings, and deliberations happen. A tribunal seated in London may hear witnesses in Dubai, take documents in Singapore, and deliberate over video call, yet the arbitration remains, in the eye of the law, a London arbitration. This article explains the distinction in plain language and then tests it against the leading authorities from England, India, and the wider #Model_Law world. It examines the statutory architecture of the #UNCITRAL_Model_Law and the #New_York_Convention, the interpretive rules courts use when parties are careless with their drafting, the long and unsettled Indian debate running from BALCO to Arif Azim and Disortho, and the recent statutory reset in England through the Arbitration Act 2025. The article closes with practical drafting guidance for students and early practitioners, and a short reflection on why the seat still matters in an age of remote hearings and talk of #delocalisation. Keywords: #seat_of_arbitration; #venue; #lex_arbitri; #party_autonomy; #New_York_Convention; #enforcement_of_awards; #supervisory_jurisdiction; #arbitration_agreement 1. Introduction International arbitration exists because parties from different legal systems rarely trust each other's national courts. A company in Kuala Lumpur does not want to be sued in New Delhi, and a company in New Delhi does not want to be dragged before a Malaysian judge. Arbitration offers a neutral private tribunal whose decision, the #arbitral_award, can be enforced across borders under a widely ratified treaty. That promise, however, rests on a network of legal connections that most commercial parties never think about until a dispute forces them to. The most important of those connections is the seat. The vocabulary is where trouble begins. Contracts speak loosely of the "place" of arbitration, the "venue" of hearings, the "location" of proceedings, and the "seat" of the reference, often in the same clause and often as though they meant the same thing. They do not. The seat is a legal address. The #venue is a street address. One determines rights, the other determines logistics. When a drafting lawyer treats them as interchangeable, a court somewhere may later spend years, and the parties may spend a fortune, deciding what the clause actually meant. This article is written for students and for practitioners in the early part of their careers. Its aim is modest but useful: to make the seat and venue distinction genuinely clear, and to show why the distinction is not academic hair-splitting but a live commercial risk. The structure is straightforward. Section 2 sets out the core concepts and the reason the terminology is so slippery. Section 3 explains what the choice of seat actually decides, from the governing #procedural_law to the enforceability of the final award. Section 4 explains what the venue decides, which is much less, and why moving hearings around does not move the seat. Section 5 turns to interpretation, examining how English courts read arbitration clauses that mention only a venue, through the well-known #Shashoua_principle. Section 6 traces the tangled Indian jurisprudence, which is the richest and most contradictory body of case law on this exact question anywhere in the common law world. Section 7 examines a closely related puzzle that students often merge with the seat question, namely the law that governs the #arbitration_agreement itself, and the English statutory reform of 2025. Section 8 offers concrete drafting advice. Section 9 places the debate in comparative and policy context. Section 10 concludes. A short warning before we begin. The seat and venue distinction touches several adjacent ideas that are easy to confuse with it: the law governing the main contract, the law governing the arbitration agreement, the procedural rules chosen by the parties, and the seat itself. Keeping these four apart is half the battle. The reader who leaves this article able to separate them will already understand the subject better than a surprising number of the litigants whose disputes fill the reports. 2. Conceptual foundations: seat, place, and venue 2.1 The seat as a legal construct The #seat_of_arbitration, sometimes called the legal seat, the juridical seat, or the locus arbitri, is not a physical spot on a map. It is a legal construct that ties a particular arbitration to a particular national legal system. When parties choose London as their seat, they are not choosing a building or even really a city. They are choosing English arbitration law to be the framework within which their arbitration lives and, if necessary, dies. This is the point made repeatedly in the leading treatises. Gary Born describes the arbitral seat as the legal or juridical home of the arbitration rather than a geographical location, and Redfern and Hunter make the same observation with their memorable image of the notional motorist: once you have chosen the seat, you must obey its mandatory arbitration laws just as a driver must obey the traffic laws of the country through which he happens to be driving, whether or not he finds them convenient. The consequences flow from that single choice. The seat supplies the #lex_arbitri, the law of the arbitration, which governs matters such as the constitution of the tribunal, the powers of the arbitrators, the availability of court assistance, the standards for challenging arbitrators, and the grounds on which the award can be attacked. The seat also fixes which country's courts have #supervisory_jurisdiction over the reference. And, crucially for enforcement, the seat gives the award its nationality. An award made in an arbitration seated in Paris is, for the purposes of the #New_York_Convention, a French award, even if every hearing took place in Geneva. 2.2 The venue as physical geography The #physical_venue, by contrast, is exactly what it sounds like: the place where the physical business of the arbitration happens. Witnesses are heard there. Counsel make their submissions there. The tribunal may deliberate there. But the venue carries no automatic legal freight. Choosing to hold a hearing in Dubai for the convenience of a Gulf-based witness does not subject the arbitration to Emirati arbitration law if the seat is elsewhere. The venue is chosen for practical reasons: cost, travel time, the availability of hearing rooms, the location of documents, the comfort and safety of participants. These are logistical concerns, not legal ones. The distinction can be stated as a single sentence that students should memorise. The seat answers the question "under which country's arbitration law and courts does this arbitration operate," while the venue answers the question "where will people physically gather." The first is about legal control. The second is about convenience. 2.3 Why the word "place" causes so much trouble Much of the confusion in this field comes from a single word: place. The #UNCITRAL_Model_Law and many national statutes use "place of arbitration" to mean what this article calls the seat. Yet in ordinary speech "place" sounds physical, and drafters frequently use it, and its cousin "venue," to mean the location of hearings. So the same word carries two meanings, and a clause that says "the place of arbitration shall be Singapore" might be read as fixing the legal seat or, less plausibly but not impossibly, as merely naming a convenient hearing location. Article 20 of the Model Law tries to defuse this. Article 20(1) allows the parties to agree on the "place of arbitration," and in the absence of agreement leaves it to the tribunal. Article 20(2) then adds that, notwithstanding the place fixed under Article 20(1), the tribunal may meet at any place it considers appropriate for hearings, deliberations, or the inspection of goods or documents. In other words, the Model Law itself draws the seat and venue line: the place fixed under 20(1) is the legal seat, and the freedom to meet elsewhere under 20(2) is the venue power. The trouble is that the drafters used the same word, "place," for the legal concept, and left "venue" and "location" floating in ordinary usage. That verbal overlap is the root of a remarkable amount of litigation. The tidiest solution, which modern reform efforts increasingly favour, is to abandon "place" in the legal sense and speak only of the "seat." The 2024 Indian expert committee chaired by T. K. Viswanathan recommended precisely that, proposing that the statutory text replace "place" with "seat" everywhere except where the physical hearing location is meant, in which case "venue" should be used. That reform, if enacted, would remove at a stroke a great deal of the ambiguity that Indian courts have wrestled with for two decades. 3. What the seat actually decides The importance of the seat is not abstract. The choice produces at least six concrete legal consequences, and it is worth walking through each so that the stakes are clear. 3.1 The procedural law, or lex arbitri The first and most direct consequence is that the seat supplies the #lex_arbitri, the mandatory procedural law of the arbitration. This is not the same as the arbitral rules the parties may have chosen, such as the ICC, LCIA, SIAC, or HKIAC rules. Those institutional rules operate within, and are subordinate to, the mandatory law of the seat. Where the two conflict on a mandatory point, the law of the seat prevails. The #lex_arbitri typically governs the appointment and removal of arbitrators, the tribunal's powers to order interim measures, the extent to which local courts may assist or intervene, the form requirements for a valid award, and the grounds on which an award may be challenged. A seat with a modern, arbitration-friendly law, minimal court interference, and experienced judges is therefore a real asset. A seat with an outdated or interventionist law is a liability that no clever drafting of the institutional rules can fully cure. 3.2 Supervisory jurisdiction of the courts The second consequence is that the courts of the seat hold #supervisory_jurisdiction over the arbitration. These are the courts that can be asked to appoint an arbitrator when the parties deadlock, to grant interim relief in support of the arbitration, to determine a challenge to an arbitrator, and, most significantly, to hear an application to set aside the award. This supervisory role is exclusive in the sense that courts elsewhere should not entertain applications to annul an award made under a foreign seat; their role is limited to recognising and enforcing, or refusing to enforce, that award. The link between seat and supervisory court is so strong that some jurisdictions treat a choice of seat as functionally equivalent to an exclusive jurisdiction clause in favour of the seat's courts, a point the Indian Supreme Court made forcefully in Indus Mobile Distribution v Datawind and reaffirmed in later cases. 3.3 The nationality of the award The third consequence concerns the nationality of the #arbitral_award. Under the #New_York_Convention, an award is treated as made in the country of the seat. This matters because the Convention's machinery for cross-border #enforcement_of_awards distinguishes between the country where the award was made, which alone may annul it, and the many other countries where it may be enforced. Article V(1)(e) of the Convention allows enforcement to be refused if the award has been set aside by a competent authority of the country in which, or under the law of which, the award was made. That phrase points squarely at the seat. So the seat is not just about how the arbitration is run; it determines which single legal system has the power to kill the award at source. 3.4 The grounds and forum for setting aside The fourth consequence follows from the third. Because only the courts of the seat may annul an award, the #setting_aside regime of the seat, its grounds, its time limits, and its judicial attitude, becomes decisive. In Model Law jurisdictions the grounds for #annulment under Article 34 are narrow and mirror the enforcement grounds under Article 36 and the Convention's Article V. In non-Model-Law jurisdictions the grounds may be broader or narrower, and the courts may be more or less willing to intervene on the merits. A losing party will look first to the courts of the seat for any chance of overturning the award, and a winning party will pray that the seat's annulment law is tight. This is one reason parties fight so hard over the seat, and over whether a mentioned location is the seat at all. 3.5 Arbitrability and public policy The fifth consequence is that the seat's law helps determine which disputes are capable of being resolved by arbitration at all, a question lawyers call #arbitrability, and what the seat regards as matters of #public_policy that cannot be contracted around. Some categories of dispute, certain competition, insolvency, or consumer matters for example, may be non-arbitrable under the seat's law, and an award purporting to resolve them may be vulnerable. Public policy at the seat can also affect the validity of the award. These are mandatory constraints that the parties cannot displace merely by choosing friendly institutional rules. 3.6 The law governing the arbitration agreement, indirectly The sixth consequence is more subtle and is examined in detail in Section 7. In many legal systems the seat has become the default source of the law governing the #arbitration_agreement itself, at least where the parties have not made an express choice. This is a distinct question from the #lex_arbitri and from the law of the main contract, but the seat increasingly supplies the answer by default. The English Arbitration Act 2025 has now made this explicit for English-seated arbitrations, as discussed below. 3.7 Court support and interim measures There is a further, practical consequence that students often overlook because it does not appear until a dispute is already under way. The courts of the seat are the natural source of court-ordered support for the arbitration. If a party needs an urgent freezing order to stop assets disappearing, an order to preserve evidence, or an order compelling a witness to attend, it will usually look to the courts of the seat, because those courts have the clearest jurisdiction to assist a reference seated within their territory. Many modern statutes, including the English Arbitration Act 1996 as amended and its equivalents elsewhere, give the seat's courts express powers to grant such relief in support of the arbitration. A seat whose courts are willing and quick to grant supportive orders is worth a great deal when a dispute turns urgent. A seat whose courts are slow, hostile, or unpredictable can leave a party without effective protection at the very moment protection matters most. The venue, once again, is irrelevant to this. It is the seat's courts, not the courts of the city where the last hearing happened, that a party will ask for help. Taken together, these consequences explain why practitioners regard the seat as the single most important choice in an arbitration clause, more important, in many cases, than the choice of institution or even the substantive governing law of the contract. The seat is the load-bearing wall of the whole structure. 4. What the venue decides, and why it does not move the seat If the seat carries all this legal weight, what does the #venue do? The honest answer is: very little, in legal terms. The venue determines where people physically meet. That is a matter of convenience, and the freedom to choose it flexibly is one of arbitration's genuine advantages over litigation. 4.1 The Model Law and institutional rules on hearing location Article 20(2) of the #UNCITRAL_Model_Law, mentioned earlier, expressly permits the tribunal to meet at any appropriate location for hearings and deliberations, regardless of the fixed place of arbitration. The major institutional rules say the same in their own words. Under such rules the seat once fixed remains constant, while hearings may be held wherever is sensible. The point is that changing the hearing location does not change the seat, and therefore does not change the #lex_arbitri, the supervisory court, or the nationality of the award. A tribunal seated in Paris that holds three weeks of hearings in London is still a Paris-seated tribunal producing a French award. English courts have no supervisory power over it merely because the witnesses gave evidence in a London hotel. 4.2 The remote hearing question The rise of remote and hybrid hearings has sharpened this point rather than blurred it. When a hearing takes place by video, with the arbitrators in three different countries and the witnesses in a fourth, there is no meaningful single physical venue at all. Far from destabilising the concept of the seat, this development highlights why the seat must be a legal construct rather than a physical one. The arbitration needs a legal home even when it has no physical one. Modern institutional rules and revised statutes make clear that holding hearings remotely, or in a country other than the seat, does not alter the seat. The award is still deemed to be made at the seat. This is a reassuring feature: the parties' careful choice of a stable, arbitration-friendly seat is not undone by the practical decision to save everyone a long flight. 4.3 The danger of confusing the two in drafting The problem, then, is not that venue and seat conflict in principle. The problem is that drafters name a venue and forget to name a seat, and a court is later asked whether the named venue was in fact intended to be the seat. This is where the tidy conceptual distinction meets the messy reality of interpretation, and it is the subject of the next two sections. The lesson for the drafter is simple and worth stating twice: name the seat expressly, name it clearly, and if you also want to fix or suggest a hearing venue, use different words and say so. Confusing "seat" with "venue" in a clause is the single most common cause of the litigation this article describes. 5. Determining the seat when the clause is unclear: the English approach Well-drafted clauses name the seat and cause no difficulty. Badly drafted clauses, which are common, force courts to work out what the parties meant. English law has developed the most influential set of interpretive tools for this task, and those tools have been borrowed across the common law world. 5.1 Naviera Amazonica and the birth of the distinction in English law The English courts drew the seat and venue line clearly in Naviera Amazonica Peruana v Compania Internacional de Seguros del Peru. The clause in that case pointed in two directions, and the Court of Appeal had to untangle the legal home of the arbitration from its physical setting. The judgment recognised that an arbitration has a juridical seat that carries the #curial_law, and that this is conceptually separate from the place where hearings might be conducted. That recognition, unremarkable now, was foundational then, and it set English law on the path of treating the seat as a legal anchor rather than a mere address. 5.2 The Shashoua principle The most cited English contribution is the #Shashoua_principle, named after Roger Shashoua v Mukesh Sharma. The clause in that case designated London as the venue of arbitration, adopted a supranational body of arbitral rules, and named no other place as the seat. The court held that in those circumstances the reference to London as the venue amounted to the choice of London as the juridical seat. The reasoning was that where a clause fixes a single venue, chooses an established set of arbitral rules to govern the procedure, and gives no contrary indication pointing to some other seat, the natural and indeed inescapable inference is that the named venue was intended to be the seat. The principle can be broken down into elements. First, there is an express designation of a place as the venue. Second, there is no separate designation of any other place as the seat. Third, the parties have chosen a supranational or institutional set of rules to govern the arbitration. Fourth, there is no significant contrary indication. When all four are present, the venue is treated as the seat. The logic is that reasonable commercial parties who name a place and pick a recognised procedural framework are choosing a legal home, not merely booking a conference room. 5.3 The strength of the seat as an exclusive jurisdiction agreement English authority also treats the agreement on a seat as analogous to an exclusive jurisdiction clause in favour of the courts of that seat, at least for matters concerning the supervision of the arbitration. This is why, once the seat is settled, the courts of other jurisdictions are expected to keep their hands off the supervisory functions. The seat is not a loose preference; it is a binding allocation of judicial authority. That strength is part of why the interpretive exercise matters so much. Deciding that London rather than New Delhi is the seat does not merely relocate the hearings. It transfers the power to appoint, to supervise, and to annul. 5.4 When the parties name nothing at all A separate problem arises when a clause is silent on both seat and venue. Here the interpretive question is not whether a named venue is the seat, but how the seat is to be identified when the parties have said nothing useful. The statutory framework supplies a fallback. Under the English Arbitration Act 1996 the seat may be determined by the parties, by any arbitral institution or person the parties have authorised, or by the tribunal if so authorised, and failing all of that, by reference to the parties' agreement and the relevant circumstances. The #UNCITRAL_Model_Law takes a similar approach in Article 20, leaving the place of arbitration to the tribunal where the parties have not agreed it, having regard to the circumstances of the case and the convenience of the parties. Older English authority such as the Dubai Islamic Bank litigation illustrates the exercise the courts undertake when no seat has been designated and one must be worked out. The important point for students is that a seat will always be found; an arbitration cannot proceed as a legal nullity floating free of every system of law. Silence does not create a delocalised arbitration. It simply hands the determination of the seat to the tribunal, the institution, or ultimately a court, and produces exactly the uncertainty that a clear clause would have avoided. 5.5 Seat, jurisdiction, and the anti-suit dimension English law has also developed the idea that an agreement on a seat behaves like a submission to the supervisory jurisdiction of the seat's courts, so much so that those courts may act to protect the arbitration. In cases such as C v D the English courts treated the choice of an English seat as carrying with it English law as the law governing the arbitral process and the parties' recourse against the award, and were willing to restrain a party from pursuing challenges in a foreign court that would be inconsistent with the parties' choice of seat. This protective function reinforces the practical weight of the seat. It is not merely a label; it can be enforced against a party who tries to escape it. The venue enjoys no such protective status, because nothing legal turns on it. 5.6 The limits of the Shashoua reasoning The Shashoua reasoning was born in a specific context. London arbitration under international rules was, as the court noted, a well-known phenomenon chosen by foreign parties precisely for its neutral and developed arbitration law. That contextual rationale does not automatically transplant to every jurisdiction. As Indian commentators have pointed out, the assumption that a named venue plus international rules signals a deliberate choice of legal seat is easiest to justify where the venue is a global arbitration hub with a mature #lex_arbitri. Applying the same presumption mechanically to a purely domestic setting, where the parties may have named a city simply because it was convenient, is more contestable. This tension, between a clean presumption and a search for genuine intention, runs right through the Indian case law examined next. 6. The Indian saga: two decades of contradiction No jurisdiction has generated more case law on the seat and venue distinction than India, and none has generated more contradiction. The story is worth telling in some detail, because it is the best available illustration of how much can turn on the distinction and how badly things go wrong when the courts themselves cannot agree. 6.1 The starting point: Bhatia International and BALCO The Indian Arbitration and Conciliation Act 1996 is modelled on the #UNCITRAL_Model_Law, and Part I of the Act, which contains the supervisory and annulment provisions, was intended to apply to arbitrations seated in India. In Bhatia International v Bulk Trading, however, the Supreme Court read Part I as capable of applying even to arbitrations seated outside India, on the ground that the relevant section did not contain the word "only." That reading exposed foreign-seated arbitrations to Indian court intervention and was widely criticised as undermining the territorial logic of the Model Law. The correction came in Bharat Aluminium Co v Kaiser Aluminium Technical Services, universally known as BALCO. A five-judge bench held that Part I of the Act applies only to arbitrations seated in India, restoring the territorial principle that the seat determines which courts supervise. BALCO also embraced the seat and venue distinction and approved the #Shashoua_principle. As a decision of a five-judge Constitution Bench, BALCO carries great authority, and every later case must be read against it. 6.2 Hardy Exploration: the concomitant factor test Trouble returned in Union of India v Hardy Exploration and Production. The arbitration clause named Kuala Lumpur as the venue and adopted the UNCITRAL rules, but did not expressly designate a seat. A three-judge bench held that a named venue does not automatically become the seat; something more, a concomitant factor showing that the parties intended the venue to be the seat, is required. On that reasoning the court declined to treat Kuala Lumpur as the seat and, controversially, allowed Indian courts a role. Critics observed that this approach sat uneasily with BALCO's endorsement of the #Shashoua_principle, under which the named venue plus international rules and no contrary indication should have made Kuala Lumpur the seat. 6.3 BGS SGS Soma: the venue is presumptively the seat A year later, in BGS SGS Soma JV v NHPC, another three-judge bench took the opposite view. The clause provided that proceedings would be held at "New Delhi/Faridabad." The court applied the #Shashoua_principle in full and held that where a clause designates a single venue, fixes the arbitration to that place, adopts a governing set of rules, and shows no significant contrary indication, the named venue is the juridical seat. On the facts, because the proceedings were conducted and the award signed in New Delhi, New Delhi was the seat, and its courts had exclusive supervisory jurisdiction. The court went further and declared that Hardy Exploration was not good law, because it had failed to follow the #Shashoua_principle approved by the Constitution Bench in BALCO. Here lies the central doctrinal knot. BGS Soma and Hardy Exploration were both three-judge benches. Under settled principles of Indian precedent, a bench cannot overrule an earlier bench of equal strength; it can only doubt it and refer the matter to a larger bench. So although BGS Soma declared Hardy Exploration wrong, it could not formally overrule it. The two decisions therefore coexist in uneasy tension, and lower courts have had to choose between them. The result is a body of law in which the answer to a fundamental question, does a named venue presumptively become the seat, depends partly on which line of authority a particular judge prefers. 6.4 Mankastu Impex: back to intention The Supreme Court added a further wrinkle in Mankastu Impex v Airvisual. The clause there did not use the words "seat" or "venue" at all. It provided that the arbitration would be administered in Hong Kong, that the place of arbitration would be Hong Kong, that Indian law governed the contract, and that the courts of New Delhi would have jurisdiction. The court declined to apply a mechanical presumption and instead searched for the parties' true intention from the clause as a whole and their conduct. It concluded that Hong Kong was the seat. The method resembled Hardy Exploration's search for intention more than BGS Soma's bright-line presumption, even though the outcome, treating the named place as the seat, resembled BGS Soma. Mankastu therefore neither resolved the conflict nor clearly took a side, and commentators noted that it deepened rather than dispelled the uncertainty. 6.5 Arif Azim and Disortho: the recent chapter The most recent decisions have leaned towards the BGS Soma line while, oddly, ignoring Hardy Exploration altogether. In Arif Azim Co v Micromax Informatics FZE, decided in late 2024, the clause named Dubai as the venue and adopted UAE arbitration rules. Applying BGS Soma, the Supreme Court held that where only one place is designated, the proceedings are fixed to that place, a curial law is stipulated, and there is no significant contrary indication, the named place is the juridical seat rather than a mere venue. Dubai was therefore the seat. The court treated the seat, once identified, as carrying exclusive supervisory jurisdiction, consistent with the analogy to an exclusive jurisdiction clause. In Disortho SAS v Meril Life Sciences, decided in 2025, the court again relied on BGS Soma and on Arif Azim, and again passed over Hardy Exploration without engaging it. Commentators have criticised this selective treatment. Because Hardy Exploration, BGS Soma, Mankastu, and Arif Azim are all decisions of coequal benches, the cleanest way to settle the field would be a reference to a larger, five-judge bench that could reconcile or choose between the competing approaches. Until that happens, the Indian position is best described as strongly favouring the presumption that a named venue is the seat, absent significant contrary indicia, while a formally unresolved contrary authority still lingers in the background. 6.6 The jurisdictional payoff and the fixed seat Two further Indian threads help complete the picture. The first concerns what the seat does once it is identified. In Indus Mobile Distribution v Datawind the Supreme Court held that the moment a seat of arbitration is fixed, it operates like an exclusive jurisdiction clause, conferring supervisory jurisdiction on the courts of the seat even where no part of the cause of action arose there. This is a striking illustration of how the seat overrides the ordinary rules of civil jurisdiction that would otherwise send a case to the courts where the contract was performed or the defendant resides. The choice of seat is, in effect, a choice of forum for everything to do with supervising and challenging the award. Earlier, in Brahmani River Pellets v Kamachi Industries, the court had already leaned towards treating a designated place as carrying exclusive jurisdiction, foreshadowing the presumption later crystallised in BGS Soma. The second thread concerns the stability of the seat once chosen. In Inox Renewables the Supreme Court accepted that a juridical seat, once fixed, can be changed only by the mutual agreement of the parties, not by the unilateral act of one side or by the mere fact that hearings drift to a different city. This is the practical counterpart of the seat and venue distinction. Hearings can move freely, because they are venue. The seat cannot move casually, because it is the legal anchor, and moving it would relocate the supervisory court, alter the #lex_arbitri, and potentially change the nationality of the award. The stability of the seat is precisely what makes it valuable, and precisely why courts guard it against accidental relocation dressed up as a change of venue. 6.7 What the Indian experience teaches The Indian saga is more than a curiosity. It teaches three durable lessons. First, the seat and venue distinction is not a theoretical nicety; it decides which courts control an arbitration and whether an award can be attacked at home. Second, sloppy drafting, naming a venue but not a seat, using "place" ambiguously, or combining a named venue with a jurisdiction clause pointing elsewhere, is the direct cause of the litigation. Third, even sophisticated apex courts can disagree profoundly about the right interpretive method, which means the only reliable protection for a commercial party is to draft the seat clearly in the first place and never leave the question to a court's later reconstruction of intention. 7. The related puzzle: the law governing the arbitration agreement Students frequently conflate the seat with a different but neighbouring question: which law governs the #arbitration_agreement itself? Because this question interacts with the seat, and because English law recently changed its answer, it deserves separate treatment. 7.1 Three different laws in one contract A single arbitration clause can implicate at least three distinct bodies of law. There is the law governing the main contract, the substantive law that decides whether a party breached its obligations. There is the #lex_arbitri, the procedural law of the seat, which governs how the arbitration is conducted and supervised. And there is the law governing the arbitration agreement, the clause to arbitrate, which decides questions such as whether that clause is valid, whom it binds, and what disputes it covers. These three can, in principle, be three different legal systems. A contract might be governed by New York law, seated in London, with an arbitration agreement governed by, say, French law. Keeping them apart is essential, and much confusion in this area comes from treating the seat's law as automatically governing the arbitration agreement, or from assuming the main contract's law does so. 7.2 Enka v Chubb and the English common law position Before 2025, English law approached the law of the #arbitration_agreement through the framework set out by the Supreme Court in Enka Insaat ve Sanayi v OOO Insurance Company Chubb. The court held, broadly, that where the parties had chosen a governing law for the main contract, that choice would generally extend to the arbitration agreement as an implied choice; and where they had not, the arbitration agreement would be governed by the law with which it was most closely connected, which would usually be the law of the seat. The practical effect was that arbitration agreements in London-seated arbitrations could end up governed by a foreign law, the law of the main contract, even though the arbitration was seated in England and English arbitration law was chosen to run it. Many practitioners found this awkward, because it exposed English-seated arbitrations to foreign rules on questions such as scope and #arbitrability, and generated satellite disputes. 7.3 The Law Commission and the Arbitration Act 2025 The Law Commission of England and Wales reviewed the Arbitration Act 1996 and, in its 2023 Final Report and draft Bill, recommended reversing the Enka default. The recommendation was that, absent an express choice by the parties, the law of the arbitration agreement should be the law of the seat. Parliament enacted this through the Arbitration Act 2025, which received Royal Assent in February 2025 and whose substantive provisions came into force on 1 August 2025. A new provision inserted into the 1996 Act now states, in effect, that the law applicable to an arbitration agreement is the law the parties expressly agree, or, where there is no such agreement, the law of the seat. Importantly, the Act clarifies that merely choosing a law for the main contract does not by itself count as an express agreement that the same law governs the arbitration agreement. Certain treaty-based and investment arbitration agreements are carved out. The effect is to strengthen the seat's role. For English-seated arbitrations, the seat now supplies not only the #lex_arbitri and the supervisory court but also, by default, the law governing the arbitration agreement. This is a deliberate policy choice aimed at certainty and at reinforcing London's standing as a leading seat against competitors such as Singapore, Hong Kong, and Paris. It also means that the older English case law giving priority to the main contract's law no longer states the default position for arbitrations to which the 2025 Act applies. 7.4 A point of comparison It is worth noting that the English reform sets the default differently from some other major seats. In Singapore and Hong Kong the prevailing approach still tends to treat the law of the main contract as the implied choice for the arbitration agreement, closer to the pre-2025 English position. This divergence is a reminder that these questions are matters of national policy, not universal truth, and that the "right" answer depends on the seat. For students the lesson is that the interaction between seat and the law of the arbitration agreement varies by jurisdiction, and that the only way to avoid the uncertainty entirely is to state the governing law of the arbitration agreement expressly, which almost no standard clause does. 8. Drafting guidance: turning theory into a safe clause Everything above points to a single practical conclusion. Most seat and venue disputes are self-inflicted wounds caused by careless drafting, and almost all of them are preventable. The following guidance distils the case law into concrete drafting habits. 8.1 A worked illustration Consider a simple hypothetical that pulls the strands together. A buyer in Mumbai and a seller in Dubai sign a supply contract. The dispute clause says: "Any dispute shall be settled by arbitration. The arbitration shall be held in Dubai. The contract is governed by the laws of India. The courts at Mumbai shall have exclusive jurisdiction." A dispute arises. The buyer wants Indian courts to supervise, arguing that Indian law governs the contract and Mumbai has exclusive jurisdiction, so Dubai is merely the #physical_venue. The seller wants the arbitration treated as Dubai-seated, arguing that Dubai is the only place named for the arbitration itself, so it is the #legal_seat. Walk through the analysis. The clause designates a single place for the arbitration, Dubai, and does not separately name any other seat. Under the #Shashoua_principle and the BGS Soma line, that points strongly towards Dubai being the seat. The choice of Indian law for the contract is a choice of substantive law, not of the #lex_arbitri, and does not by itself make India the seat. The Mumbai jurisdiction clause is the real troublemaker: it is the kind of contrary indication that a court might, or might not, treat as significant. A judge following the presumption would likely hold Dubai to be the seat and read the Mumbai clause as confined to non-arbitration court proceedings. A judge searching harder for intention might treat the Mumbai clause as evidence that the parties wanted Indian supervision, and reach the opposite result. That is precisely the uncertainty the drafters created, and it could take years and two levels of appeal to resolve. Had the clause simply said "the seat of arbitration is Dubai; hearings may be held in Mumbai or elsewhere for convenience," there would have been nothing to argue about. The worked example shows that the risk is not exotic. It is built into ordinary, plausible-looking clauses that thousands of contracts contain. 8.2 Drafting habits First, name the seat expressly and use the word "seat." A clause that says "the seat of arbitration shall be London, England" leaves nothing for a court to reconstruct. Avoid relying on the word "place," which carries a double meaning, and avoid relying on "venue" to do the seat's work. Second, if a hearing location matters to the parties, deal with it separately and use different language. For example, a clause might say that the seat is London while adding that hearings may be held in a named city or wherever the tribunal considers convenient. This makes explicit that the venue is a matter of logistics and not a covert designation of the seat. Third, do not undercut the seat with an inconsistent jurisdiction clause. A common trap is to name a seat in the arbitration clause while a general jurisdiction clause elsewhere in the contract confers exclusive jurisdiction on the courts of a different country. Because a choice of seat operates like an exclusive jurisdiction agreement for supervisory matters, an inconsistent general clause invites exactly the kind of argument that fills the Indian reports. Reconcile the two clauses, or make clear that the general jurisdiction clause does not apply to arbitration-related court proceedings. Fourth, consider stating the law governing the #arbitration_agreement expressly. Given the divergence between jurisdictions on the default rule, and the recent English reversal, an express choice removes a whole category of potential disputes. If the parties are content for the seat's law to govern the arbitration agreement, say so; if they want the main contract's law to govern it, say that instead. Fifth, choose the seat for its law and its courts, not for its hotels. The attractiveness of a seat lies in its #lex_arbitri, the quality and neutrality of its supervisory courts, its ratification of the #New_York_Convention, and its track record of enforcing awards and resisting improper interference. These are the features that will matter if the dispute turns nasty. The comfort of the hearing rooms, by contrast, is a venue question that can be solved later and cheaply. Sixth, when using institutional rules, understand how those rules fix or default the seat. Some rules allow the institution to determine the seat if the parties have not, and the outcome may not be what the parties would have chosen. Naming the seat expressly overrides the default and keeps control with the parties, which is the whole point of #party_autonomy. A clause that follows these six habits will rarely, if ever, generate a seat and venue dispute. A clause that ignores them may generate years of litigation over a question the parties could have answered in a single sentence. 9. Comparative and policy perspectives 9.1 The competition among seats Behind the doctrine lies a commercial reality: seats compete. London, Singapore, Hong Kong, Paris, Geneva, and, increasingly, Dubai through the DIFC and other hubs, market themselves as reliable homes for international arbitration. The English reforms of 2025 were explicitly framed as an effort to keep London attractive against rivals that had modernised their laws more recently. Singapore and Hong Kong have built reputations for efficient, arbitration-friendly courts and modern statutes. This competition benefits users, because it pushes seats towards clearer laws, faster court support, and reliable #enforcement_of_awards. But it also means that the choice of seat is a strategic decision with real consequences, and that a party who chooses a weak or interventionist seat may regret it long after the hearing venue has been forgotten. 9.2 The delocalisation debate A recurring theoretical question is whether an arbitration must have a seat at all, or whether it can float free of any national legal system. This is the #delocalisation debate. Advocates of delocalisation argue that international arbitration derives its authority from the parties' agreement and the international enforcement regime, not from the law of any particular country, and that the arbitration should therefore be freed from the supervisory reach of the seat's courts. Critics, and the dominant practice, insist that an award needs a legal home to have legal force, and that the #New_York_Convention itself presupposes a country in which the award is made. The prevailing view, reflected in the treatises and in the case law of leading jurisdictions, is firmly that arbitrations are seated, that the seat carries mandatory law, and that a genuinely anational, delocalised arbitration remains more a theoretical possibility than a practical reality outside a few special regimes. The rise of remote hearings, as noted earlier, does not support delocalisation; it simply separates the physical venue from the legal seat even more visibly than before, while leaving the seat firmly in place. 9.3 The enforcement lens It helps to view the whole subject through the lens of enforcement, because that is where the stakes become concrete. A winning party wants an award that can be enforced widely and annulled nowhere except, at most, at a friendly seat. The #New_York_Convention makes the seat central to this calculation. It ties the power of annulment to the seat, and it allows enforcing courts elsewhere to refuse enforcement on the limited grounds in Article V, including where the award has been set aside at the seat, where the #arbitration_agreement was invalid under the law to which the parties subjected it or, failing that, the law of the seat, and where the composition of the tribunal or the procedure was not in accordance with the parties' agreement or the law of the seat. Every one of these grounds points back, directly or indirectly, to the seat. The venue appears nowhere in this machinery. That asymmetry captures the entire lesson of this article in a single observation: the enforcement regime cares intensely about the seat and not at all about where the hearings happened. 9.4 The scale of what is at stake It is easy to treat this subject as a lawyer's puzzle, but the numbers show why states legislate so carefully around it. Surveys of arbitration users have repeatedly ranked London among the most preferred seats in the world, with Singapore, Hong Kong, Paris, and Geneva close behind, and the Law Commission's review of the English legislation estimated that there are at least several thousand domestic and international arbitrations seated in England each year, worth a very large sum to the national economy in arbitrator and legal fees alone. A jurisdiction that captures a reputation as a reliable seat attracts not just fees but the surrounding ecosystem of counsel, experts, institutions, and hearing services. This is why the seat has become an object of policy competition and why reforms are framed in the language of national economic advantage. For the individual party, the same forces translate into a straightforward incentive: choosing a well-regarded seat means choosing a system that has been deliberately tuned to make arbitration work, to support it through the courts, and to enforce its awards. The venue, by contrast, generates hearing-room bookings and hotel nights. Both matter to somebody, but only one of them decides who wins the fight over the award. Another practical dimension is timing. Disputes about the seat tend to erupt at the worst possible moment, after an award has been made, when the losing party is hunting for any procedural foothold to resist enforcement or to seek annulment. A party that discovers only at that stage that its clause is ambiguous about the seat is in a poor position. The arguments are expensive, the delay is real, and the outcome is uncertain because, as the Indian line of cases shows, even senior courts can reasonably disagree. The cost of prevention, a clear sentence at the drafting stage, is trivial by comparison. This asymmetry between the tiny cost of clarity and the large cost of ambiguity is the strongest possible argument for taking the distinction seriously. 9.5 The direction of reform The direction of travel across jurisdictions is towards clarity and towards reinforcing the seat's role. England's 2025 statute made the seat the default source of the arbitration agreement's governing law. India's 2024 expert committee recommended purging the ambiguous word "place" from the statute in favour of "seat" and "venue" used precisely. Institutional rules increasingly spell out that hearings elsewhere do not disturb the seat and that awards are deemed made at the seat. The common thread is a recognition that the seat and venue distinction is too important to be left to accident, and that legislatures and institutions can do a great deal to prevent disputes by insisting on clear terminology. The remaining gap, as the Indian case law shows, is judicial consistency where the underlying drafting is poor and the statute is ambiguous, and even that gap narrows as reform proceeds. 10. Conclusion The distinction between the seat and the venue is one of those foundational ideas that looks trivial and turns out to be decisive. The #legal_seat is the juridical home of an arbitration. It supplies the #lex_arbitri, fixes the courts with #supervisory_jurisdiction, gives the award its nationality, controls #setting_aside and, in a growing number of jurisdictions including England after 2025, supplies the default law of the #arbitration_agreement. The #physical_venue supplies nothing of legal weight. It is the geography of convenience, and it can move from city to city, or dissolve into a video call, without disturbing any of the legal consequences that flow from the seat. The case law surveyed here, from the English foundations in Naviera Amazonica and the #Shashoua_principle to the long and contradictory Indian line from BALCO through Hardy Exploration, BGS Soma, Mankastu, Arif Azim, and Disortho, all illustrates the same point from different angles. Where parties name the seat clearly, there is no dispute. Where they name only a venue, or use "place" ambiguously, or contradict themselves with a stray jurisdiction clause, courts are forced to reconstruct intention, and the results are unpredictable and expensive. The statutory reforms in England and the reform proposals in India both aim to shrink this zone of ambiguity by insisting on clear language. For students and new practitioners the practical message is the most valuable takeaway. Understand that the seat and the venue are different in kind, not merely in degree. Understand that the seat is chosen for its law and its courts, and the venue for its convenience. And understand that a single well-drafted sentence naming the seat expressly will save a client from the years of litigation that fill the reports of every major arbitration jurisdiction. The distinction between the seat and the venue is not a trap for the unwary so much as a reward for the careful. Master it, and much of #international_arbitration falls into place. Hashtags #seat_vs_venue #arbitration_law #international_arbitration #legal_seat #physical_venue #lex_arbitri #New_York_Convention #UNCITRAL_Model_Law #party_autonomy #supervisory_jurisdiction #arbitration_agreement #enforcement_of_awards #Shashoua_principle #Arbitration_Act_2025 #dispute_resolution References Books and treatises Blackaby, N., Partasides, C., Redfern, A., and Hunter, M. (2023) Redfern and Hunter on International Arbitration. 7th edn. Oxford: Oxford University Press. Born, G. B. (2021) International Commercial Arbitration. 3rd edn. Alphen aan den Rijn: Kluwer Law International. Born, G. B. (2021) International Arbitration: Law and Practice. 4th edn. Alphen aan den Rijn: Kluwer Law International. Reports and reference works Law Commission of England and Wales (2023) Review of the Arbitration Act 1996: Final Report and Bill (Law Com No 413, HC 1787). London: Law Commission. Expert Committee on Arbitration Law (Chair: T. K. Viswanathan) (2024) Report of the Expert Committee to Examine the Working of the Arbitration Law and Recommend Reforms in the Arbitration and Conciliation Act 1996. New Delhi: Ministry of Finance, Government of India. The Cambridge Compendium of International Commercial and Investment Arbitration (2023) "Lex Arbitri and the Rules of Procedure". Cambridge: Cambridge University Press. DOI: https://doi.org/10.1017/9781108304467.018 Max Planck Encyclopedia of International Procedural Law (2024) "Seat of Arbitration". Oxford: Oxford University Press. Articles and commentary Kluwer Arbitration Blog (2025) "Hardy Forgotten, Doctrines Entangled: Disortho v. Meril Life and the Seat-Venue Conundrum". Alphen aan den Rijn: Wolters Kluwer. Legislation and international instruments Arbitration Act 1996 (United Kingdom). Arbitration Act 2025 (United Kingdom). Arbitration and Conciliation Act 1996 (India). Convention on the Recognition and Enforcement of Foreign Arbitral Awards 1958 (New York Convention). UNCITRAL Model Law on International Commercial Arbitration 1985 (as amended in 2006). Cases Arif Azim Co Ltd v Micromax Informatics FZE (2024), Supreme Court of India. Bharat Aluminium Co v Kaiser Aluminium Technical Services Inc (2012) 9 SCC 552, Supreme Court of India (BALCO). Bhatia International v Bulk Trading SA (2002) 4 SCC 105, Supreme Court of India. BGS SGS Soma JV v NHPC Ltd (2020) 4 SCC 234, Supreme Court of India. Disortho SAS v Meril Life Sciences Pvt Ltd (2025), Supreme Court of India. Enka Insaat ve Sanayi AS v OOO Insurance Company Chubb [2020] UKSC 38. Indus Mobile Distribution Pvt Ltd v Datawind Innovations Pvt Ltd (2017) 7 SCC 678, Supreme Court of India. Mankastu Impex Pvt Ltd v Airvisual Ltd (2020) 5 SCC 399, Supreme Court of India. Naviera Amazonica Peruana SA v Compania Internacional de Seguros del Peru [1988] 1 Lloyd's Rep 116 (CA). Roger Shashoua v Mukesh Sharma [2009] EWHC 957 (Comm). Union of India v Hardy Exploration and Production (India) Inc (2019) 13 SCC 472, Supreme Court of India.

This article examines two features of the #common_law tradition that are often treated as opposites but in practice depend on each other: the reliance on #judicial_precedent and the capacity of judges to keep the law flexible. It explains how the doctrine of #stare_decisis works, how courts follow, distinguish, and sometimes overrule earlier decisions, and why this mixture of stability and adaptability has helped #English_common_law spread far beyond England and Wales. The study then connects these doctrinal points to a practical question that matters to businesses everywhere, namely why so many #international_commercial_contracting parties, who may have no link to the United Kingdom, still choose English law to govern their agreements and London as a place to settle disputes. Drawing on recent scholarship and on data published by arbitration institutions and legal bodies between 2021 and 2025, the article argues that #legal_certainty produced by precedent, combined with the flexibility judges retain when facts change, gives English law a distinctive commercial appeal. It also considers the main criticisms of the system and the growing competition from civil law jurisdictions and from newer commercial courts in Singapore, Dubai, and elsewhere. The article is written for students and aims to make a technical subject clear without losing accuracy. Keywords: common law; judicial precedent; stare decisis; international commercial contracts; governing law; party autonomy; legal certainty; commercial arbitration 1. Introduction Few ideas in law travel as widely as the notion that a court should follow what earlier courts have already decided. This simple habit, repeated over centuries, sits at the heart of the #common_law tradition. It is the reason a lawyer in Lagos, Sydney, Toronto, Singapore, or Dubai can read an English judgment from decades ago and still find it useful. It is also, less obviously, the reason the same tradition manages to change and adapt when the world around it changes. The relationship between fixed rules and flexible reasoning is not a weakness of the system. It is the engine that keeps it alive. The purpose of this article is to explain that relationship and to show why it matters far beyond the courtroom. The argument develops in two connected parts. The first part describes how #judicial_precedent operates: what it means to say a decision is binding, how judges identify the part of a case that must be followed, and how they retain room to move when a strict rule would produce an unfair or outdated result. The second part turns to the commercial world and asks a practical question. Why do so many companies, banks, traders, and investors, many of whom have no factory, office, or bank account in the United Kingdom, still decide that their contracts should be governed by English law and that their disputes should be resolved in London? These two questions are linked. The commercial popularity of #English_common_law cannot be understood without first understanding how precedent works, because the very thing that businesses value, namely #predictability, is produced by the discipline of following past decisions. At the same time, businesses also value the ability of the law to respond sensibly to new kinds of transactions, from complex financial products to digital assets. A system that was only rigid would be safe but stale. A system that was only flexible would be responsive but unreliable. The common law tries to hold both qualities together, and it is this balance that gives it a global reach (McKendrick, 2025). The article proceeds as follows. Section two sets out the historical and conceptual foundations of the tradition. Section three explains the doctrine of precedent in detail. Section four examines the tension between certainty and flexibility and shows the techniques judges use to manage it. Section five explains why English law is chosen so often in cross-border commerce. Section six presents recent data on governing law and #dispute_resolution. Section seven looks closely at how precedent and flexibility appear in the interpretation of commercial contracts. Section eight considers criticisms and competitors. Section nine looks at future directions, including technology and new commercial courts. Section ten concludes. 2. Historical and conceptual foundations of the common law The phrase "common law" has more than one meaning, and confusion between them is a common source of error for students. In its broadest sense, the term describes a whole legal family, the group of legal systems that trace their methods and much of their substance back to the law of England. In a narrower sense, it describes judge-made law as opposed to statute. In a narrower sense still, it refers to the body of rules developed by the old common law courts as distinct from the rules of equity that grew up alongside them. This article mostly uses the first and second meanings, treating the common law as a tradition of #case_law built up by courts over time. The tradition began to take shape after the Norman conquest of 1066, when royal judges travelled the country and gradually built a body of law that was "common" to the whole realm rather than local and varied. Over centuries these judges decided disputes one at a time, and their decisions accumulated into a store of reasoning that later judges consulted and respected. The system was never designed in advance by a single lawmaker. It grew, case by case, out of the practical need to decide real disputes and to decide like cases alike. This bottom-up quality distinguishes it sharply from the #civil_law systems of continental Europe, which are organised around comprehensive written codes drafted by legislators and scholars (Siems, 2022). The contrast should not be overstated. Modern common law jurisdictions are full of statutes, and modern civil law jurisdictions pay close attention to earlier decisions. Comparative scholars increasingly warn against treating the two families as sealed boxes and encourage a more global view of how legal ideas move and mix (Pirie, 2024; Siems and Yap, 2024). Even so, a real difference in method survives. In the common law world, the reported decision of a court is itself a source of law. A judge does not merely apply the code and illustrate it with an example; the judge makes law when deciding the case, and that law binds others. In the civil law world, the primary source is the code, and decisions, however influential, are formally treated as applications of the code rather than as independent law. This methodological difference has a direct consequence for how each system grows. Because the common law develops through decided cases, it tends to grow incrementally, in small steps tied closely to concrete facts. Judges are cautious about laying down broad rules that reach beyond the dispute in front of them, because they know their words may bind future courts. This caution is itself a source of #legal_certainty, since parties can usually predict how a court will treat a new dispute by looking at how earlier courts treated similar ones. The tradition therefore prizes analogy, close reading of facts, and careful distinctions rather than sweeping abstract statements. Two further features of the tradition deserve mention at the outset. The first is the central role of the judge. In the common law, judges are not anonymous civil servants applying a code; they are named authors whose reasoning is published, studied, criticised, and cited. The quality and independence of the judiciary therefore matter enormously to the reputation of the system, a point that becomes important later when we ask why commercial parties trust English courts. The second is the adversarial style of proceedings, in which the parties and their lawyers, rather than the judge, take the lead in gathering evidence and framing arguments. These features shape the texture of common law reasoning and help explain its appeal to sophisticated commercial users who want to control how their case is presented. A word is also needed about equity, since students often meet the term without a clear explanation. Alongside the old common law courts there grew up a separate system administered by the Court of Chancery, which developed to soften the harshness that a strict application of common law rules could sometimes produce. Equity introduced remedies and doctrines, such as the injunction, the trust, and the idea of acting in good conscience, that the common law courts did not provide. The two systems were eventually merged so that the same courts now apply both, but the distinction survives in the vocabulary of the law and in the character of many remedies. Equity is itself an early example of the tradition building in a form of flexibility to correct the rigidity of its own rules, a theme that runs through this article. 2.1 The global spread of the English common law The reach of the tradition is not an accident of history alone; it is the result of a long process by which English legal ideas were carried to, and then adopted by, much of the world. During the period of the British Empire, English law was transplanted, in varying degrees, to territories across North America, the Caribbean, Africa, South Asia, Southeast Asia, and the Pacific. When these territories became independent states, most of them kept the common law method they had inherited, adapting it to local conditions rather than discarding it. Today a large group of countries, including the United States, Canada, Australia, New Zealand, India, Nigeria, Ghana, Kenya, Malaysia, and Singapore, share the basic features of the tradition: reliance on #case_law, respect for #judicial_precedent, an independent judiciary, and adversarial procedure. This shared inheritance is what allows lawyers across the common law family to speak, in effect, a common professional language. A principle developed in one jurisdiction can be cited as #persuasive_precedent in another, and important decisions of the highest English courts have long been studied and followed well beyond England. The movement of doctrines from one system to another is sometimes described as #legal_transplant, and the common law world offers one of the clearest examples of transplantation on a global scale (Siems, 2022). The family is not uniform, and each member has developed its own distinctive rules, but the underlying method remains recognisably the same. The most important divergence within the family is that of the United States, which inherited the common law method but developed it in its own direction. American courts have generally treated precedent more flexibly than English courts, feeling freer to reconsider their own past decisions, and American law is shaped by a written federal constitution and by the coexistence of many state legal systems (Symeonides, 2021). These differences matter in practice. In global finance, New York law and English law are the two dominant choices, and although they share a common ancestry, they are not identical, and sophisticated parties choose between them deliberately. The existence of this large and successful sibling system is itself evidence of the adaptability of the common law method, which has proved able to support very different societies while keeping its essential character. The global spread of the tradition also helps explain the theme that dominates the second half of this article. Because so many commercial actors, lawyers, and judges around the world are already familiar with the common law way of reasoning, choosing #English_common_law to govern a cross-border deal does not feel foreign to them. The tradition has become a shared resource, and its reach in commerce rests in part on the reach it achieved through history. 3. The doctrine of precedent: stare decisis explained At the centre of the tradition lies the doctrine usually described by the Latin phrase #stare_decisis, which can be translated as "to stand by things decided." The doctrine holds that a court should follow the reasoning of earlier decisions when the same point of law arises again. Its purpose is to make the law consistent and foreseeable so that people can plan their affairs, so that like cases are treated alike, and so that the outcome of a dispute does not depend on the private opinions of the particular judge who happens to hear it (Lewis, 2021). It is important to be precise about what part of an earlier case is actually binding. Not every sentence in a judgment carries authority. The binding element is the #ratio_decidendi, meaning the legal principle on which the decision rests, the reasoning that was necessary to reach the result on the facts. Everything else a judge says, however interesting, is #obiter_dicta, meaning remarks made "by the way." Obiter statements can be persuasive, and famous obiter remarks have later grown into settled law, but they do not bind future courts in the way the ratio does. Much of the skill of a common law lawyer lies in reading a judgment carefully to identify the true ratio and to separate it from the surrounding commentary. Precedent operates within a #court_hierarchy. A decision binds only if it comes from a court high enough in the structure to bind the court now deciding the case. In England and Wales the hierarchy runs, in broad terms, from the magistrates' and county courts at the base, through the High Court and the Court of Appeal, up to the Supreme Court at the top. A decision of the Supreme Court binds every court below it. A decision of the Court of Appeal binds the High Court and lower courts and, subject to limited exceptions, binds the Court of Appeal itself. This vertical structure is what makes #binding_precedent possible, because it settles which decisions must be obeyed and which are merely worth considering. As the Supreme Court has observed, without such a hierarchy the law would lose its coherence and predictability (see the discussion in Willers v Joyce [2016] UKSC 44). This leads to a useful distinction between two kinds of precedent. A #binding_precedent is one that a court is obliged to follow because it comes from a superior court in the same hierarchy on the same point of law. A #persuasive_precedent is one that a court may consider and may choose to follow but is not obliged to follow. Persuasive authority includes decisions of lower courts, decisions of courts in other common law countries, decisions of the Judicial Committee of the Privy Council, and the reasoning of respected scholars. The willingness of English courts to treat decisions from Australia, Canada, Singapore, and elsewhere as persuasive is one of the quiet ways the common law family stays connected across borders, and it helps explain why the tradition retains a shared character despite being spread across many independent states. The strictness of the doctrine has varied over time and place. The doctrine is applied more rigidly in England than in the United States, where courts have historically felt freer to reconsider their own past decisions (Symeonides, 2021). Within England, the high point of strictness came with the decision in London Tramways Co Ltd v London County Council [1898] AC 375, in which the House of Lords held that it was absolutely bound by its own previous decisions. For most of the twentieth century, the highest court in the land could not correct its own mistakes; only Parliament could. That rigidity created stability but also the risk that an error, once made at the top, would harden into permanent law. As the following section shows, the tradition eventually found ways to soften this rigidity without abandoning the discipline that gives precedent its value. A final point about the doctrine concerns its underlying justification. Following precedent is not merely a matter of habit or convenience. It reflects deeper values associated with the #rule_of_law, in particular the idea that the law should treat people equally and should be knowable in advance. When a court follows precedent, it signals that the outcome does not turn on who the judge is or on how the judge feels about the parties. It curbs the personal preferences of individual judges and delivers a kind of equal treatment, since two people in the same situation receive the same answer from the law. This is why precedent is often described not just as a technique but as a commitment. 4. The tension between certainty and flexibility If precedent were absolutely rigid, the law would eventually become unjust. Society changes, technology changes, and commercial practice changes, and a rule that made perfect sense in one era can become harmful in another. A purely rigid system would also entrench mistakes, since a wrong decision could never be corrected by the courts. Yet if precedent were freely ignored, the law would lose the very #predictability that makes it useful. The genius of the common law lies in the tools it has developed to manage this tension, allowing #legal_flexibility while preserving the discipline of following past decisions. This section describes the main tools. The first and most everyday tool is #distinguishing. When a judge is faced with a precedent that would lead to an unwelcome result, the judge may find a relevant difference between the facts of the earlier case and the facts of the present one. If the facts are materially different, the earlier ratio does not apply, and the judge is free to decide the new case differently. Distinguishing is not a trick; it is a legitimate part of common law reasoning, because precedent binds only where the material facts are alike. It allows the law to develop in small, careful steps, adapting old principles to new situations without openly rejecting them. A large part of the daily life of the common law consists of arguments about whether a case is truly like an earlier one or can be distinguished from it. The second tool is #overruling. A higher court may decide that an earlier decision, whether its own or that of a lower court, was wrong and should no longer be followed. Overruling is more dramatic than distinguishing, because it removes a precedent from the law altogether. It is used sparingly, since it disturbs settled expectations and implies that earlier judges erred, but it is a necessary safety valve. The most important development in this area was the #Practice_Statement issued by the House of Lords in 1966. In that short but powerful announcement, the House declared that while it would treat its former decisions as normally binding, it would depart from a previous decision "when it appears right to do so." The statement recognised that too rigid an adherence to precedent could lead to injustice in a particular case and could unduly restrict the proper development of the law. This power now belongs to the Supreme Court, which inherited it when it replaced the House of Lords as the final court (Lewis, 2021). The Supreme Court, and the House of Lords before it, has used this power with caution. A clear example is Knauer v Ministry of Justice [2016] UKSC 9, in which the court departed from earlier authority on how to calculate certain damages, holding that the older approach produced results that were out of step with reality and with developments in the law. The court stressed that it would not depart from precedent merely because it might have decided the earlier case differently; there had to be a good reason, such as a change in circumstances or a demonstrable defect in the earlier reasoning. This measured approach is important. The value of the departure power lies precisely in its restrained use, because a court that overruled its own decisions freely would reintroduce the uncertainty the doctrine is meant to prevent. The Court of Appeal has its own, narrower, escape routes. Under the rule established in Young v Bristol Aeroplane Co Ltd [1944] KB 718, the Court of Appeal is generally bound by its own previous decisions but may depart from them in a small number of defined situations, for example where two of its own past decisions conflict, or where an earlier decision cannot stand with a later decision of the Supreme Court, or where a decision was reached carelessly and overlooked a relevant statute or binding authority. These exceptions are deliberately limited, so that the intermediate appellate court cannot casually unsettle the law but can still correct obvious problems. A third and more subtle mechanism is the incremental development of legal principle. Because the common law grows case by case, judges are constantly refining and extending existing rules to fit new facts. They rarely announce a revolution. Instead they build outward from settled points, treating the law as a structure that can be added to without being demolished. This incrementalism is a form of flexibility, but it is a disciplined flexibility, tethered at every step to what has gone before. It allows the law of contract, tort, and commerce to keep pace with changing practice while remaining recognisably continuous with its past. Taken together, these tools show that certainty and flexibility are not really enemies in the common law. They are two sides of the same practice. The discipline of following precedent gives the law stability, while distinguishing, overruling, and incremental development give it the capacity to change. A skilled judge uses all of these tools to keep the law both reliable and just. It is this combination, rather than either quality alone, that commercial users find attractive, as the next section explains. 5. Why English common law dominates international commercial contracting We can now turn to the practical heart of the article. When two companies from different countries make a deal, they must decide which country's law will govern their contract and where any dispute will be resolved. These choices are made through a governing law clause and a jurisdiction or arbitration clause. They are among the most important terms in any cross-border agreement, because the same words in a contract can be read very differently under different legal systems, and because the choice of forum can decide how quickly, how fairly, and at what cost a dispute is resolved. The freedom of the parties to make these choices is known as #party_autonomy, and it is now accepted as a near universal principle of private international law (Girsberger, Kadner Graziano and Neels, 2021). The remarkable fact is how often parties with no connection to the United Kingdom choose #English_common_law. English law is frequently described as a kind of global public utility, selected to govern transactions that have little or nothing to do with Britain, simply because it is predictable and commercial. Several features of the system explain this appeal, and most of them flow directly from the doctrine of precedent discussed above. The first feature is #predictability produced by a deep body of #case_law. Because English courts have decided a very large number of commercial disputes over a very long time, and because those decisions are published and binding, there is an unusually rich store of authority on almost every commercial question. A lawyer advising on an English law contract can usually find a decided case on point and can predict with reasonable confidence how a court would rule. This depth of precedent lowers legal risk, and lower legal risk can affect how a deal is priced or even whether it goes ahead at all. Predictability is not an abstract virtue here; it has direct commercial value. The second feature is respect for #freedom_of_contract. English law generally takes the view that commercial parties are the best judges of their own interests and should be held to the bargains they make. It is reluctant to rewrite contracts in the name of fairness, reluctant to imply terms that the parties did not agree, and reluctant to police the substance of a deal between sophisticated businesses. Crucially, English law does not recognise a broad, overarching duty of good faith in the performance of ordinary commercial contracts, unlike many civil law systems and unlike the law of some other jurisdictions. For a business, this means the contract it signed is, in most cases, the contract it will get, without unexpected obligations being read in later. Many commercial parties see this predictability of outcome as a major advantage (McKendrick, 2024). The third feature is the quality and independence of the judiciary and the commercial bar. English commercial judges are typically drawn from experienced practitioners and are widely respected for their expertise and impartiality. The courts are seen as independent of government and of the parties, they do not generally award punitive damages, and damages are set by judges rather than by juries. For a foreign company suing or being sued far from home, the perceived neutrality and competence of the tribunal is a powerful reassurance. This reputation is inseparable from the culture of reasoned, published judgments that the doctrine of precedent requires. The fourth feature is the flexibility discussed in the previous section. A legal system that could not adapt would be a poor choice for fast-moving commerce. English law has repeatedly shown that it can develop its principles to accommodate new kinds of transactions, from modern financial instruments to novel forms of security, without discarding the certainty that makes it attractive. Commercial parties therefore get the best of both worlds: a stable and well-mapped body of law, together with courts capable of applying that law sensibly to situations the earlier cases never imagined. A fifth feature, easy to overlook, is language and networks. English is the dominant language of international business, and English law comes with a vast professional ecosystem of law firms, barristers, arbitrators, textbooks, and precedents written in that language. Once a system reaches this scale, it tends to reinforce itself, because using the most widely understood law and language reduces friction, and because each new deal governed by English law adds to the store of expertise and authority that makes the next choice easier. This self-reinforcing quality helps explain why the position of English law in #cross_border_trade has proved so durable. A sixth factor is a matter of caution rather than enthusiasm. Commercial lawyers repeatedly advise that a party should never agree to be governed by a law it does not understand, because an unfamiliar system may contain surprises, such as mandatory rules, good faith requirements, or short limitation periods, that change the meaning and value of the deal. Faced with a choice between the counterparty's home law and their own, parties often settle on a neutral third system that both sides can respect and that both sides' advisers can competently handle. English law is frequently that neutral option. It is well understood internationally, its content is documented in an enormous professional literature, and neither side needs to feel it is submitting to the other's home advantage. This role as a trusted neutral ground is one of the least glamorous but most powerful reasons for its selection. The choice of #governing_law is also closely tied to the choice of forum. Parties usually try to match the two, so that the court or tribunal deciding the dispute is one that knows the chosen law well. If a contract is governed by English law, the parties will often also choose the English courts or a London-seated arbitration, because those forums are best placed to apply English law correctly. This matching effect reinforces the position of English law and English forums together, since a decision to use one tends to pull the other along with it. It is another example of how the different attractions of the system combine and support each other rather than operating in isolation. None of this means English law is objectively superior for every purpose, and the later sections consider its critics. But it does explain the pattern observed in practice. The combination of predictable precedent, respect for the parties' bargain, a trusted judiciary, and disciplined flexibility gives #international_commercial_contracting parties strong reasons to select English law and English forums, and the data examined next show how strong that pattern is. 6. The empirical picture: governing law and dispute resolution Claims about the popularity of English law are not merely impressions. They are supported by data published each year by courts, arbitration institutions, and professional bodies. Read with appropriate care, since some of these bodies have an interest in promoting their own jurisdictions, the figures still paint a consistent picture. Consider international #commercial_arbitration, the preferred method for resolving many cross-border disputes because arbitral awards are easier to enforce across borders than court judgments under the widely ratified New York Convention (Blackaby, Partasides and Redfern, 2022). The London Court of International Arbitration reports that London remains the most popular seat for the arbitrations it administers, that the overwhelming majority of its cases involve parties from outside the United Kingdom, and that English law is by a wide margin the most frequently chosen #governing_law, selected in more than eight out of ten of its cases in recent years. The great majority of its caseload is international in character, which shows that parties are choosing English law and a London seat for reasons unconnected to any British nationality of the parties. The pattern extends well beyond London-based institutions. The International Chamber of Commerce, based in Paris, has reported that English law was the single most common governing law chosen in its new cases in a recent year, ahead of every other national law, even though parties selected the laws of well over one hundred different jurisdictions across its caseload. The Singapore International Arbitration Centre has reported that English law was the second most common choice of governing law in its cases, behind only Singapore law, and English law has also ranked among the most common choices in cases administered by institutions in Hong Kong and Stockholm. Taken together, these figures show that the appeal of English law is genuinely global and is not confined to any one arbitral institution or region. Litigation before the courts tells a similar story. The Commercial Court in London handles a large volume of disputes each year, and a substantial majority of the cases it decides are international in nature, meaning the subject matter or the parties have no necessary connection to the United Kingdom. Analysts have estimated that a very large share of international financial contracts and long-term trade agreements contain clauses in favour of English or New York governing law, reflecting the dominance of these two systems in global finance and trade. The London court is often described as delivering more written commercial judgments each year than comparable courts elsewhere, and it has a strong record of resolving many contested trials quickly. A particular stronghold of English law is the maritime and shipping sector. More shipping disputes are referred to London than to anywhere else in the world, and the overwhelming majority of them are governed by English law, with a specialist body of arbitrators handling a very large number of new cases each year. This reflects both the historical role of London as a centre of trade and insurance and the depth of English case law on carriage of goods, charterparties, and marine insurance. In such specialised fields, the accumulated store of #precedent is not merely convenient; it is close to indispensable, because it answers detailed technical questions that a general code could not easily address. One reason arbitration is so central to this story is enforcement. A court judgment can be difficult to enforce in a foreign country, whereas an arbitral award can be enforced in the many states that have joined the New York Convention on the recognition and enforcement of foreign arbitral awards. This treaty framework means that a London-seated award applying English law can usually be enforced against assets almost anywhere in the world, which is a decisive advantage for parties trading across borders (Blackaby, Partasides and Redfern, 2022). The popularity of English law and a London seat therefore rests not only on the quality of the law but also on the practical assurance that a favourable result can be turned into real recovery. The economic significance of all this is considerable. Professional bodies estimate that English law governs transactions worth trillions of pounds every year across areas such as derivatives, commodities, insurance, shipping, and corporate transactions. The export value of United Kingdom legal services has continued to rise, which is taken as evidence of growing international demand for expertise in English law. These figures should be read with some caution, since the bodies that publish them, including national law societies and the arbitration institutions themselves, have an interest in presenting their jurisdictions favourably. Even after allowing for that, the consistency of the pattern across many independent sources is striking. English law functions as an infrastructure for a large part of world commerce, and that role rests on the reputation for certainty and fairness that the tradition of precedent has built. It is worth noting that the framework supporting this position is not frozen. In 2025 the United Kingdom modernised its arbitration legislation through the Arbitration Act 2025, which received Royal Assent in February 2025 and whose substantive provisions came into force in August 2025. The reform followed a careful, multi-year review by the Law Commission and was intended to keep the law up to date and to reinforce London's standing as a leading seat, rather than to overturn the existing system. Among other changes, it clarified the law governing the arbitration agreement itself, codified the duty of arbitrators to disclose matters affecting their impartiality, strengthened arbitrator immunity, and gave tribunals a clear power to dispose of hopeless claims quickly. The episode is a good illustration of the wider theme of this article. A mature legal system preserves what works while adjusting the details, combining continuity with measured change. 7. Precedent and flexibility in commercial contract interpretation To see how precedent and flexibility operate together in practice, it helps to look closely at one area that matters greatly to business, namely how English courts read and enforce commercial contracts. The case law here shows the common law method at its most characteristic, developing and refining its own principles through a sequence of decided cases (McKendrick, 2024). Take the interpretation of contract terms. English courts ask what a reasonable person, having all the background knowledge available to the parties, would have understood the words to mean. In Rainy Sky SA v Kookmin Bank [2011] UKSC 50, the Supreme Court emphasised that where words are capable of more than one meaning, the court may prefer the interpretation that makes the most commercial sense. A few years later, in Arnold v Britton [2015] UKSC 36, the court struck a more cautious note, stressing that clear language should not be displaced simply because the result seems commercially unattractive, since it is not the court's job to save a party from a bad bargain. The apparent difference between these decisions was then reconciled in Wood v Capita Insurance Services Ltd [2017] UKSC 24, where the court described interpretation as an iterative process that weighs both the language chosen and the commercial context, without giving fixed priority to either. This sequence is a textbook example of incremental development. No decision was overruled; instead the principle was refined step by step until a workable balance emerged. The law on implied terms shows the same disciplined caution. In Marks and Spencer plc v BNP Paribas Securities Services Trust Company (Jersey) Ltd [2015] UKSC 72, the Supreme Court reaffirmed that a term will be implied into a commercial contract only where it is necessary to give the contract business efficacy or is so obvious that it goes without saying. The court will not imply a term merely because it would be reasonable or fair to do so. This restraint reflects the respect for #freedom_of_contract discussed earlier. It gives parties confidence that the obligations they will face are, in the main, the ones they actually agreed, and it prevents courts from rewriting bargains after the fact. The treatment of clauses that fix a sum payable on breach offers a further illustration, this time of the law's ability to modernise an old rule. For a long time such clauses were tested by a rigid distinction between a genuine pre-estimate of loss, which was enforceable, and a penalty, which was not. In Cavendish Square Holding BV v Talal El Makdessi [2015] UKSC 67, decided together with ParkingEye Ltd v Beavis, the Supreme Court reformulated the rule. It held that the real question is whether the clause imposes a detriment on the defaulting party that is out of all proportion to any legitimate interest of the innocent party in enforcing the primary obligation. This reformulation kept the protective purpose of the old rule while making it more sensitive to modern commercial reality. It is a clear case of the law adapting a settled doctrine to serve its underlying aim more sensibly, without abandoning the certainty that commercial parties rely on. Even the vexed question of good faith fits the pattern. English law has traditionally refused to recognise a general duty of good faith in the performance of commercial contracts, preferring specific rules to a broad open-ended standard. In Yam Seng Pte Ltd v International Trade Corporation Ltd [2013] EWHC 111 (QB), a first instance judge suggested that a duty of good faith might be implied in certain long-term, cooperative contracts, sometimes called relational contracts, where the parties depend on mutual trust over time. Later cases have treated this idea with caution and have declined to turn it into a general principle. The episode shows the tradition working through a new idea slowly and carefully, testing it against the values of certainty and freedom of contract before deciding how far, if at all, it should go. Whether English law should move closer to the good faith standards common in civil law systems remains a live debate among scholars (Dagan and Peari, 2023). For those who draft and negotiate contracts, this body of case law has a very practical value. Because the courts have addressed so many recurring problems, and because their answers are binding and published, drafters can anticipate how a given clause is likely to be read and can word their agreements accordingly. A well-advised party can, in effect, borrow the accumulated wisdom of decades of litigation without having to litigate at all. This is the everyday face of #legal_certainty. It is not an abstract promise but a concrete resource that reduces the cost and risk of doing business. At the same time, the willingness of the courts to refine their approach, as the interpretation cases show, reassures drafters that the law will not apply a rigid formula blindly where the commercial context points the other way. The two qualities that this article has emphasised, certainty and flexibility, are therefore not just features of high theory; they are felt in the drafting of ordinary commercial documents. Interpretation of the contract's choice of law provisions has also generated important recent authority. In Enka Insaat ve Sanayi AS v OOO Insurance Company Chubb [2020] UKSC 38, the Supreme Court addressed the difficult question of which law governs an arbitration agreement when the parties have not said so expressly. The decision set out a structured approach, and its complexity was one reason Parliament later stepped in with a clearer default rule in the Arbitration Act 2025. This interaction between judge-made law and statute is a reminder that in a modern common law system the two sources work together. Courts develop principles through cases, and the legislature intervenes when a clearer or simpler rule is needed. The result is a layered system in which #precedent and statute each play a part. 8. Challenges, criticisms, and competitors The picture painted so far is largely positive, but a fair account must also consider the criticisms of the common law method and the growing competition from other systems. The dominance of English law is real, but it is neither guaranteed nor beyond challenge. One line of criticism concerns accessibility and cost. Because the law is spread across thousands of decided cases rather than gathered in a single code, finding the relevant rule can require expert help and expensive research. A civil law code, by contrast, aims to state the law in an organised and accessible form that an educated reader can consult directly. Critics argue that the case-based method makes the common law less transparent to ordinary people and businesses and more dependent on a professional class of lawyers. Defenders reply that codes are often more abstract than they appear and that real certainty comes from how a system is actually applied, not from the form in which its rules are written (Siems, 2022). A second criticism concerns the risk of rigidity and the slow pace of change. Because judges are reluctant to depart from precedent, an unsatisfactory rule can persist for years until a suitable case reaches a high enough court, or until the legislature acts. The strictness of the doctrine, especially before the 1966 Practice Statement, has been blamed for entrenching outdated positions. The counter-argument, developed earlier in this article, is that the tools of distinguishing, overruling, and incremental development give the system enough flexibility to correct itself, and that stability is worth a measure of delay. A third criticism is more philosophical. Some scholars question whether judge-made law is truly democratic, since judges are not elected and yet make rules that bind everyone. The response is that judges do not make law freely; they are constrained by precedent, by statute, and by the discipline of reasoned justification, and their decisions are open to correction by higher courts and by Parliament. On this view, the doctrine of precedent is itself a democratic safeguard, because it limits judicial discretion and ties each decision to what has gone before, serving the values of the #rule_of_law (Lewis, 2021). Beyond these internal criticisms lies external competition, and this is where the future of English law's dominance will be decided. Civil law jurisdictions have worked to make their systems more attractive to international business, including by setting up specialised commercial chambers and by allowing proceedings to be conducted in English. Newer centres have emerged to compete directly for cross-border work. Singapore has built a strong reputation as a seat for arbitration and has established the Singapore International Commercial Court to attract international disputes. In the Gulf, the Dubai International Financial Centre and the Abu Dhabi Global Market operate courts that apply common law principles in English and have drawn a growing caseload, and reforms in the wider United Arab Emirates have allowed some proceedings to be conducted in English. New York law remains the main rival to English law in global finance. These developments mean that parties now have real alternatives, and that English law can no longer take its position for granted. The civil law world has also mounted a more direct intellectual challenge. Some continental jurisdictions have promoted their own contract law as clearer, more accessible, and more protective of weaker parties than English law, and international bodies have produced model rules and principles intended to offer a neutral alternative to any single national system. Supporters of codified systems argue that a well-drafted code delivers certainty in a more transparent form than a mass of decided cases, and that the good faith duties common in civil law prevent the harsh outcomes that strict enforcement of a bargain can sometimes cause. These are serious arguments, and the honest response is that neither family is simply better than the other; each embodies a different balance between the freedom of the parties and the protective role of the law (Dagan and Peari, 2023; Siems, 2022). What can be said is that, for high-value commercial dealings between sophisticated parties, many users continue to prefer the predictability and the respect for the bargain that #English_common_law offers, while acknowledging that other systems may suit other purposes better. Brexit added a further complication. When the United Kingdom left the European Union, some feared that English judgments would become harder to enforce across Europe and that London would lose commercial work as a result. In practice, the choice of English law as a governing law was largely unaffected, because the rules that respect the parties' choice of law were carried over into United Kingdom domestic law, and because arbitration, which depends on the New York Convention rather than on European rules, was essentially insulated from the change. The main practical concern has been about the enforcement of court judgments within the European Union, where the position became less straightforward than before. The long-term commercial effect is still being assessed, but the early evidence suggests that the fundamentals that make English law attractive, its predictability and its flexibility, have survived the transition. The competition, however, is intensifying, and complacency would be a mistake. 9. Future directions Looking ahead, several developments are likely to test the ability of the common law to combine certainty with flexibility. The way the tradition responds will shape its global standing in the decades to come. The first is technology. Digital assets, tokenised securities, and other novel financial instruments raise questions that the older cases never contemplated. So do so-called smart contracts, where terms are recorded and performed automatically by computer code. English law has begun to grapple with these questions, and its case-based method is well suited to the task, because it can develop existing principles to fit new facts one step at a time rather than waiting for a comprehensive new code. The challenge will be to provide enough certainty for a fast-moving industry while leaving room to adjust as the technology matures. This is precisely the kind of problem the tradition has handled before, and its record of incremental development gives some reason for confidence. The second is artificial intelligence in legal work itself. Tools that can search and analyse large volumes of case law may change how lawyers and judges identify precedents and predict outcomes. If such tools make the vast body of common law more accessible and easier to navigate, they could reduce one of the main criticisms of the system, namely its cost and complexity. At the same time, they raise new questions about accuracy, accountability, and the proper role of human judgment, which the courts and the profession will need to address carefully. The third is the growing importance of values such as sustainability and responsible business conduct in commercial dealings. As contracts increasingly include obligations relating to environmental and social standards, courts will be asked to interpret and enforce terms of a kind that were rare in earlier eras. Here the debate over good faith and cooperation in long-term relationships may become more pressing, and English law will have to decide how far its traditional reluctance to imply broad duties can be maintained in the face of changing commercial expectations (Dagan and Peari, 2023). The fourth is institutional competition and cooperation. The spread of commercial courts that apply common law principles in English, from Singapore to the Gulf, may in one sense dilute the unique position of London. But it also extends the reach of the common law method and strengthens the wider family of which English law is a part. The movement of legal ideas across borders, sometimes described as #legal_transplant, has always been a feature of the tradition, and a more networked and global legal world may end up reinforcing the common law's influence even as it disperses it (Pirie, 2024; Siems and Yap, 2024). The continued modernisation of English arbitration and litigation, exemplified by the Arbitration Act 2025, suggests that the jurisdiction intends to compete actively rather than rely on past prestige. 10. Conclusion This article set out to explain two features of the common law that are usually presented as opposites, its reliance on #judicial_precedent and its capacity for flexibility, and to connect them to the global popularity of English law in commercial dealings. The central argument has been that these features are not opposites at all. The discipline of following past decisions produces the #legal_certainty that commercial parties prize, while the techniques of distinguishing, overruling, and incremental development give the law the adaptability it needs to remain just and useful as circumstances change. Certainty and flexibility are partners, not rivals. That partnership explains a great deal about the reach of #English_common_law. Companies with no link to Britain choose English law to govern their contracts and London to resolve their disputes because they can predict how the law will treat them, because the courts respect the bargains that sophisticated parties strike, because the judiciary is trusted, and because the system has repeatedly shown that it can accommodate new kinds of commerce without losing its stability. The data published by arbitration institutions and legal bodies confirm that this is not a marginal preference but a dominant pattern in #international_commercial_contracting, even as competitors in Singapore, the Gulf, New York, and continental Europe grow stronger. For students, the lesson runs deeper than the study of a single jurisdiction. The common law offers a model of how a legal system can be both reliable and responsive, how respect for the past can coexist with openness to the future, and how a tradition built one case at a time can nonetheless serve the needs of a global economy. Understanding how #precedent works, and how judges keep it flexible, is therefore not only a matter of legal technique. It is a way of understanding how law can be stable enough to trust and supple enough to last. Hashtags #Common_Law #Judicial_Precedent #Stare_Decisis #English_Common_Law #International_Commercial_Contracting #Governing_Law #Party_Autonomy #Legal_Certainty #Commercial_Arbitration #Contract_Law #Precedent_And_Flexibility #Cross_Border_Trade #London_Arbitration #Rule_Of_Law #Comparative_Law References Blackaby, N., Partasides, C., & Redfern, A. (2022). Redfern and Hunter on International Arbitration (7th ed.). Oxford University Press. https://doi.org/10.1093/law/9780192869906.001.0001 Dagan, H., & Peari, S. (2023). Choice of law meets private law theory. Oxford Journal of Legal Studies, 43(3), 520-545. https://doi.org/10.1093/ojls/gqad008 Girsberger, D., Kadner Graziano, T., & Neels, J. L. (Eds.). (2021). Choice of Law in International Commercial Contracts. Oxford University Press. Lewis, S. (2021). Precedent and the rule of law. Oxford Journal of Legal Studies, 41(4), 873-898. https://doi.org/10.1093/ojls/gqab007 McKendrick, E. (2024). Contract Law: Text, Cases and Materials (11th ed.). Oxford University Press. McKendrick, E. (2025). Contract Law (16th ed.). Hart Publishing. Peari, S., & Khanderia, S. (2021). Party autonomy in the choice of law: Some insights from Australia. Liverpool Law Review, 42(2), 275-296. https://doi.org/10.1007/s10991-021-09278-z Pirie, F. (2024). Global comparative law? Oxford Journal of Legal Studies, 45(2), 506-524. https://doi.org/10.1093/ojls/gqae038 Siems, M. (2022). Comparative Law (3rd ed.). Cambridge University Press. https://doi.org/10.1017/9781108892766 Siems, M., & Yap, P. J. (Eds.). (2024). The Cambridge Handbook of Comparative Law. Cambridge University Press. Symeonides, S. C. (2021). Choice of law in the American courts in 2020: Thirty-fourth annual survey. The American Journal of Comparative Law, 69(2), 177-262. https://doi.org/10.1093/ajcl/avab002

This article examines the origins, structure, and continuing relevance of the 1965 Convention on the Settlement of Investment Disputes between States and Nationals of Other States, widely known as the Washington Convention. The Convention created the International Centre for Settlement of Investment Disputes, the specialised institution that has become the central forum for resolving disputes between foreign investors and host governments. The study traces why the treaty was drafted in the decade of decolonisation, how its drafters designed a self-contained procedural system detached from national courts, and how its core provisions on consent, jurisdiction, applicable law, annulment, and enforcement operate in practice. It then considers the empirical reality of the modern caseload, the criticisms that have generated a sustained debate about the legitimacy of investor-state arbitration, and the reform programme now unfolding through the 2022 procedural rules, the work of the United Nations Commission on International Trade Law, and proposals for structural change. The article argues that the Convention succeeded because it solved a practical problem of trust, but that its future depends on whether the system can answer questions of fairness, consistency, and public accountability without losing the neutrality that gave it value in the first place. It is written for students who are meeting the subject for the first time and who need a clear, connected picture of the framework that stands behind almost every modern investor-state case. Keywords: international investment law; investor-state arbitration; the Washington Convention; institutional design; enforcement of awards; reform of dispute settlement. Introduction When a private company from one country invests in another, it takes a risk that goes beyond ordinary business. The company places factories, pipelines, power plants, mines, or money inside the territory of a sovereign government. That government makes the laws, runs the courts, collects the taxes, and controls the police. If a dispute later arises, the investor may find that the only place to complain is the very state it is complaining about. This is the basic imbalance that the #Washington_Convention of 1965 set out to correct. The treaty, formally the #Convention_on_the_Settlement_of_Investment_Disputes between States and Nationals of Other States, created a neutral home for these quarrels and gave it a name that students now use as shorthand for the whole system: #ICSID, the International Centre for Settlement of Investment Disputes. The Convention is often described as the backbone treaty of #investor_state_arbitration, and the description is accurate. Almost every well-known investment case of the last three decades has been decided either under the ICSID Convention itself or under rules that ICSID administers. When a tribunal orders a state to pay hundreds of millions of dollars to a mining company, or when a government successfully defends a claim that its environmental regulation was really a disguised taking of property, the procedural spine of that case usually runs back to a treaty signed in Washington in 1965. To understand modern #foreign_investment law, a student must first understand this document and the institution it built. This article explains the framework from the ground up. It begins with the historical setting that made the treaty possible, because the Convention cannot be understood apart from the political moment that produced it. It then works through the architecture of the institution and the operational core of the treaty: how a tribunal gets #jurisdiction, what law it applies, how its decisions can be challenged, and how a winning party actually collects. The later parts turn to the living system. They set out what the caseload actually looks like today, examine the serious criticisms that have gathered around the process, and describe the reform effort that is reshaping the field. The aim throughout is to keep the language plain while respecting the real complexity of the subject. The Historical Background: Why the Convention Was Needed 2.1 The old world of gunboats and diplomacy Before the Convention existed, an investor who was mistreated abroad had two main options, and neither worked well. The first was to sue in the courts of the #host_State. This was rarely attractive, because a foreign company suing a government in that government's own courts faced obvious doubts about neutrality. The second option was to ask its own government to take up the claim on its behalf, a practice known as #diplomatic_protection. Under this older model, the investor's home state would press the host state through diplomatic channels, and in the most extreme historical cases through economic pressure or even force. This model created two problems. It turned commercial disagreements into confrontations between governments, which meant an ordinary contract dispute could poison relations between whole nations. It also made the investor dependent on political convenience, because a home government would only take up a claim when it suited its wider foreign policy. A small firm with a legitimate grievance might be ignored, while a large firm with political connections might receive help it did not deserve. The result was a system that was both unstable and unfair. 2.2 The pressures of the mid-twentieth century By the 1950s and 1960s the world was changing quickly. #Decolonisation was creating dozens of newly independent states across Asia, Africa, and the Caribbean, and these states wanted foreign capital to build their economies. At the same time, several waves of nationalisation, in which governments took over foreign-owned industries, had frightened investors and made capital cautious. Newly independent governments were suspicious of arrangements that looked like a continuation of colonial control, while investors were suspicious of legal systems they did not know and could not predict. The tension produced a stalemate. #Capital_importing_states needed investment but resisted anything that felt like a surrender of sovereignty. Capital-exporting states and their companies wanted protection but could not impose it. What both sides lacked was a trusted middle ground: a mechanism that a government could accept without feeling humiliated and that an investor could rely on without feeling exposed. 2.3 The role of the World Bank and Aron Broches The institution that broke the deadlock was the #World_Bank. As a development body owned by governments from every part of the world, the Bank had credibility on both sides of the divide. It could speak to capital importers as a partner in development and to capital exporters as a guardian of financial confidence. The key intellectual figure was #Aron_Broches, the Bank's General Counsel, who is rightly regarded as the architect of the Convention. Broches understood that the project could only succeed if it was modest in ambition and neutral in design. He deliberately avoided writing substantive rules about what states could or could not do to investors. The Convention says almost nothing about #expropriation, compensation, or fair treatment. Instead it does one thing extremely well: it builds a procedure. This choice, sometimes called the procedural focus of the Convention, was a stroke of diplomatic genius. States could sign a treaty about how disputes would be handled without committing themselves in advance to any particular standard of behaviour. The substance would come later, mostly through thousands of separate investment treaties that would use ICSID as their preferred forum. The Convention was adopted on 18 March 1965 and entered into force on 14 October 1966 after the required number of ratifications. It has since attracted very broad participation. By the mid-2020s more than one hundred and sixty states had signed the treaty and around one hundred and fifty-eight had ratified it, making it one of the most widely accepted instruments in international economic law. The central promise it offered was #depoliticization: the removal of investment disputes from the arena of state-to-state confrontation and their placement inside a calm, rule-based process. 2.4 The genius of leaving substance out It is worth dwelling on the drafting choice that made everything else possible, because it is easy to miss how unusual it was. Most treaties try to lay down rules of conduct. They say what parties must do and must not do. The Convention refuses to do this. The document that accompanied the treaty, the Report of the Executive Directors of the World Bank, was explicit that the Convention did not aim to state any rules of substantive law. It neither guaranteed the investor a right to compensation nor obliged the state to protect property in any particular way. It only offered a road, not a destination. This restraint had two effects that shaped the whole future of the field. First, it made the treaty easy to accept. A government could join without conceding anything about how it would treat investors, because the treaty said nothing on that subject. Second, and less obviously, it created an empty space that had to be filled from somewhere else, and that somewhere else turned out to be the network of investment treaties that states signed with one another in the following decades. Each of those treaties supplied the substantive standards that the Convention had deliberately omitted, and most of them named ICSID as the forum in which those standards would be enforced. In this way the 1965 treaty and the later treaty network fit together like a lock and a key. The Convention built the courtroom; the investment treaties wrote the law that would be argued inside it. Neither would have achieved much alone, but together they produced the most active system of international adjudication involving private parties that the world has ever seen. This is the deeper meaning of the claim that the Convention is the backbone of the field: it is the load-bearing structure on which a whole body of #international_investment_law came to rest. The Architecture of the Institution 3.1 A part of the World Bank family ICSID is one of the five organisations that make up the World Bank Group, and it is based at the Bank's headquarters in Washington. This connection is important for practical reasons. It gives the Centre administrative stability, financial backing, and a reputation for impartiality that a stand-alone body might struggle to build. At the same time, ICSID is not a court in the ordinary sense. It does not decide cases itself. It is better understood as an administering body, a kind of specialised secretariat that organises and supports arbitrations conducted by independent tribunals. 3.2 The governing organs Three organs structure the institution. The first is the #Administrative_Council, on which every contracting state is represented, usually by its World Bank governor. The Council is the policy-making body. It adopts the procedural rules, approves the budget, and elects the head of the Centre. Crucially, it operates on the principle that each member has one vote, which reinforces the sense that the system belongs equally to capital importers and capital exporters. The second organ is the Secretariat, headed by the #Secretary_General. The Secretariat performs the day-to-day work: it receives requests for arbitration, keeps the register of cases, assists tribunals, manages hearings, and holds the funds. One quietly powerful role of the Secretary-General is the screening function. A request to start a case will not be registered if the dispute is manifestly outside the jurisdiction of the Centre. This is a narrow gate, designed only to filter out cases that are obviously ineligible, but it protects states from being dragged into proceedings that plainly do not belong at ICSID. The third element is the Panels. Each contracting state may designate individuals to a Panel of Arbitrators and a Panel of Conciliators. These panels are a resource from which appointments can be drawn, particularly when the parties cannot agree, although in practice most #arbitral_tribunal members are chosen by the parties from outside the panels. 3.3 A procedure, not a court It is worth pausing on the difference between ICSID and a permanent court such as the International Court of Justice. A permanent court has standing judges who hear every case. ICSID has no standing judges. For each dispute a fresh tribunal is assembled, usually of three members, and it dissolves once the case is over. This design gives parties considerable control over who decides their dispute, which many find reassuring. It also, as later parts of this article explain, produces one of the deepest structural criticisms of the system, because different tribunals can reach different conclusions on the same legal question with no higher body to reconcile them. Jurisdiction: The Heart of the System 4.1 Why consent is everything No tribunal can hear a case unless it has authority to do so, and under the Convention that authority rests entirely on #consent. This is the single most important idea in the whole framework. A state is never forced to appear before an ICSID tribunal simply because it has joined the Convention. Joining the treaty only makes the machinery available. The state must separately agree to use that machinery for a particular category of disputes. Article 25 of the Convention captures this in a phrase that students should memorise: the jurisdiction of the Centre extends to disputes that the parties have consented in writing to submit. Consent can be given in several ways, and understanding them explains how the modern system actually functions. The oldest route is a clause in a contract between the investor and the state, in which both sides agree that any dispute will go to ICSID. A second route is a provision in a national investment law, by which a state offers arbitration to all qualifying foreign investors. By far the most important modern route, however, is the investment treaty. 4.2 The treaty route and the idea of arbitration without privity In the decades after 1965, states signed thousands of #bilateral_investment_treaty agreements and a smaller number of multilateral instruments. Many of these treaties contain a clause in which each state promises, in advance and to the world, to arbitrate disputes with investors from the other state. When an investor later files a claim, it accepts that standing offer. The two acts together, the state's general promise and the investor's later acceptance, form the agreement to arbitrate. This mechanism is sometimes described as arbitration without privity, because the state and the investor never signed a single document together. Their consent is assembled from two separate sources across time. The consequence is enormous. It means that a private company can bring a sovereign government before an international tribunal on the strength of a treaty the government signed with a third country, perhaps years earlier and without any specific dispute in mind. According to ICSID's own statistics, treaties remain the dominant basis of consent, with bilateral investment treaties historically underpinning the majority of registered cases. The #Energy_Charter_Treaty and regional agreements such as the former #NAFTA framework have also generated significant numbers of claims. A simple worked example makes the point clearer. Imagine that Country A and Country B sign a bilateral investment treaty in which each promises to arbitrate at ICSID any dispute with an investor from the other. Ten years later a company incorporated in Country A builds a power plant in Country B. Country B then cancels the plant's licence without paying compensation. The company can now file a claim at ICSID even though it never negotiated any arbitration clause with Country B and even though the treaty was signed before the company's investment existed. The company's decision to file is treated as its acceptance of Country B's standing offer, and the agreement to arbitrate springs into being at that moment. This is how the vast majority of modern cases begin. It also explains why questions of corporate nationality can decide the fate of a claim: if the company can show it is truly a national of Country A, the treaty applies; if a tribunal decides the company is only a shell used to gain access to a favourable treaty, jurisdiction may be refused. The generosity of the consent mechanism and the strictness of the nationality inquiry are two sides of the same coin. 4.3 The two dimensions of jurisdiction Jurisdiction under Article 25 has two dimensions, and both must be satisfied. The first concerns the parties, sometimes called jurisdiction over persons. On one side there must be a contracting state or a designated subdivision or agency of that state. On the other side there must be a national of another contracting state, which may be an individual or a company. Questions of nationality can become surprisingly complex, especially where a company is owned through a chain of holding entities registered in different countries, a practice that critics label #treaty_shopping. The second dimension concerns the subject matter, and here the central and much-debated word is #investment. The Convention deliberately does not define investment. The drafters left the term open so that it could adapt to changing commercial reality. This silence has forced tribunals to develop their own tests. 4.4 The meaning of investment and the Salini test The most influential attempt to give the word content is the so-called #Salini_test, named after an early case involving a construction project. Under this approach a transaction is more likely to count as an investment if it involves a contribution of money or assets, a certain duration, an element of risk, and a contribution to the economic development of the host state. Not every tribunal treats all four elements as strict requirements, and some regard the last one, contribution to development, as too vague to be a real condition. The debate continues, but the test remains a standard reference point. For a student, the key lesson is that the boundary of the system is not fixed by the treaty text; it is shaped case by case by the tribunals themselves, which is a recurring feature of this field. The Arbitral Process and the Applicable Law 5.1 Building the tribunal Once a case is registered, the parties turn to constituting the tribunal. The usual model is three members: one appointed by each side and a president chosen jointly or, if the parties cannot agree, by the Chairman of the Administrative Council. The Convention requires arbitrators to be persons of high moral character and recognised competence, and it limits the ability of the parties to fill a tribunal entirely with their own nationals. These safeguards are meant to protect the neutrality of the panel, although as later sections show, the way arbitrators are selected has itself become a subject of concern. Tribunals enjoy the power to decide on their own jurisdiction, a principle familiar throughout arbitration. They may also grant #provisional_measures, which are temporary orders designed to preserve the position of the parties while the case proceeds, for example by asking a party not to take steps that would aggravate the dispute. These powers give the process practical teeth during the long months a case can take. 5.2 The law the tribunal applies A tribunal must decide the dispute according to some body of law, and the Convention addresses this in a provision that students should know well: #Article_42. It provides that the tribunal shall decide the dispute in accordance with the rules of law agreed by the parties. In the absence of such agreement, the tribunal applies the law of the host state together with the applicable rules of international law. This dual reference to national and international law is significant. It means a host state cannot escape its international obligations simply by pointing to its own domestic legislation. If a national law conflicts with a binding rule of international law, the international rule can prevail. This is one of the ways the Convention protects investors against the risk that a government might change its own laws to defeat a legitimate claim. In treaty cases the picture is often simpler, because the investment treaty itself supplies the substantive standards, such as protection against unlawful expropriation and the guarantee of #fair_and_equitable_treatment, and the tribunal applies those standards directly. 5.3 The award The end product of the process is the #award. The Convention requires the award to deal with every question submitted and to state the reasons on which it is based. A tribunal member who disagrees may attach a separate or dissenting opinion. The award is written, final in the sense described in the next section, and binding on the parties. Because the whole system is built to produce an enforceable result, the drafting and reasoning of the award carry great weight; a poorly reasoned award is not only unpersuasive but may be vulnerable to challenge. A Self-Contained System: Annulment Instead of Appeal 6.1 No appeal to national courts One of the most distinctive and deliberate features of the Convention is that it creates a #self_contained_system. An ICSID award cannot be reviewed by the courts of any country. There is no appeal to a national judge on the ground that the tribunal got the law wrong or misjudged the facts. This is very different from most other forms of arbitration, where the losing party can usually ask a local court to set the award aside. The drafters insulated ICSID awards from national courts precisely to remove the fear that a host state's own judiciary might undermine an unfavourable result. 6.2 The narrow ground of annulment The Convention does not leave the losing party entirely without a remedy, but it channels that remedy through a single internal route: #annulment under #Article_52. Annulment is not an appeal. An appeal asks whether the decision was correct; annulment asks only whether the process was fundamentally sound. The grounds are limited and closed. An award may be annulled if the tribunal was not properly constituted, if it manifestly exceeded its powers, if there was corruption on the part of a member, if there was a serious departure from a fundamental rule of procedure, or if the award failed to state the reasons on which it was based. A request for annulment is decided not by an ordinary court but by an #ad_hoc_committee of three persons appointed for that purpose. If the committee annuls the award, in whole or in part, the dispute can be resubmitted to a new tribunal. The important point is that annulment is meant to be exceptional. It exists to catch serious breakdowns in the integrity of the process, not to give disappointed parties a second attempt at winning on the merits. 6.3 The consistency problem The absence of an appeal has a cost, and it is one of the central themes of the modern reform debate. Because there is no higher body to declare what the law is, different tribunals can and do interpret similar treaty language in different ways. Two investors in comparable situations may receive opposite answers, and there is no formal mechanism to reconcile the contradiction. Annulment cannot solve this, because a committee cannot annul an award simply for being legally wrong. The result is a system that prizes finality and independence but sometimes struggles to deliver predictability, a tension that runs through the criticisms discussed later. Recognition and Enforcement of Awards 7.1 The enforcement promise An award is only useful if it can be turned into real money or real action. The Convention delivers on this through one of its strongest provisions, #Article_54. Every contracting state agrees to recognise an ICSID award as binding and to enforce the financial obligations it imposes as if the award were a final judgment of that state's own highest courts. This is a remarkable commitment. It means an investor holding an ICSID award can, in principle, go to the courts of any of the many contracting states, present the award, and have it treated almost automatically as an enforceable judgment. This enforcement route is often contrasted with the one available for ordinary commercial arbitration awards under the #New_York_Convention. Under that older treaty, a national court asked to enforce an award may refuse on several grounds, including public policy. Under Article 54 of the ICSID Convention, the review is far narrower. National courts are not supposed to reopen the merits or second-guess the tribunal. Their role is close to ministerial: to register the award and let enforcement proceed. This gives ICSID awards a strength that few other international decisions enjoy. 7.2 The limit of sovereign immunity There is, however, an important limit. The Convention distinguishes between recognition and enforcement on one hand and actual execution against assets on the other. Article 55 preserves the law of #sovereign_immunity as it applies to execution. In plain terms, an investor may obtain full recognition of an award, but when it tries to seize a state's property to satisfy that award, it can still run into the rule that certain sovereign assets, such as diplomatic premises or military property, are protected from seizure. This is why a winning investor sometimes holds a valid award for years without collecting, because the state has no unprotected assets in a convenient location. 7.3 Recent enforcement litigation The relationship between enforcement and immunity has produced significant litigation in the 2020s, particularly in England. In a series of cases arising from awards against European states, English courts examined whether a state could resist the registration of an ICSID award by pleading immunity. The English Court of Appeal held in 2024 that by agreeing to Article 54 the contracting states had, in effect, submitted to enforcement in one another's courts, so that immunity could not be used to block registration. The matter reached the United Kingdom Supreme Court, which in 2026 confirmed that ratifying the Convention amounts to a submission to the enforcement jurisdiction of the courts of other contracting states, drawing a firm line between recognition, which immunity cannot defeat, and execution against assets, which remains governed by the ordinary immunity rules. These decisions matter because they reinforce the practical value of an ICSID award and align the English position with courts in Australia, the United States, France, and elsewhere. They show the enforcement architecture of the 1965 treaty being tested and largely upheld sixty years after it was written. The Additional Facility The Convention as originally written only covered disputes where both the state and the investor's home state were members of ICSID. This left gaps. What about an investor whose home country had not joined, or a dispute that did not quite fit the definition of investment? To fill these gaps, ICSID created the #Additional_Facility in 1978. The Additional Facility is a set of rules administered by the Centre that allows it to handle certain cases falling outside the Convention itself. Awards rendered under the Additional Facility do not enjoy the automatic enforcement guarantee of Article 54, because that guarantee flows from the Convention. Instead they are enforced through the ordinary channels available to commercial awards. In 2022 the Additional Facility rules were widened so that proceedings can take place even where neither party is a member state or the national of one, and so that a regional economic organisation such as the #European_Union can be a party. This last change reflects the reality that investment agreements are increasingly signed by regional blocs rather than individual countries. The Additional Facility shows the flexibility of the ICSID system: rather than amending the Convention, which is very hard to do, the Centre expanded its usefulness through a separate set of rules. 8A. Beyond Arbitration: Other Ways to Settle a Dispute Although arbitration dominates the caseload, the Convention was never meant to offer only one path. From the beginning it provided for conciliation, a gentler process in which a neutral commission helps the parties reach a voluntary agreement rather than imposing a binding decision. Conciliation has always been used far less often than arbitration, mainly because parties who reach the point of a serious dispute usually want a decision they can enforce rather than a recommendation they can ignore. Even so, its presence in the treaty shows that the drafters saw dispute resolution as a spectrum, not a single door. In recent years the Centre has widened that spectrum. As part of the modernisation completed in 2022, ICSID introduced dedicated rules for #mediation and for fact-finding. Mediation is a flexible, confidential process led by a neutral third party who helps the sides talk toward a #settlement, and it can be used on its own or in parallel with an arbitration. Fact-finding allows a neutral expert to establish disputed facts before positions harden. These tools respond to a growing sense that many investment disputes are better solved by negotiation than by a long and costly fight, especially where the investor and the state expect to keep working together for years to come. There is one further point that students often miss. ICSID does not only administer cases under its own rules. The Secretariat also acts as an administering authority for a number of investor-state cases conducted under other procedures, in particular the arbitration rules of the United Nations Commission on International Trade Law. This means the Centre's practical role in the field is even larger than its own caseload suggests, because it supports proceedings that are not, strictly speaking, ICSID cases at all. The institution built by the 1965 treaty has therefore become a general service provider for #dispute_resolution in the investment field, not merely the keeper of a single set of rules. ICSID in Practice: The Modern Caseload 9.1 From a slow start to the central forum For its first two decades ICSID was almost idle. The first case was not registered until 1972, and for many years only a handful of disputes reached the Centre. The transformation came with the spread of investment treaties in the 1990s. As thousands of treaties came into force, each containing a standing offer to arbitrate, the number of claims rose sharply. By the 2020s ICSID had become, by a wide margin, the busiest institution of its kind in the world. The scale is now considerable. According to ICSID's published statistics, by the end of 2024 the Centre had registered more than one thousand cases under the Convention and the Additional Facility since it began, and it crossed the milestone of its thousandth case that year. In the 2024 calendar year alone around fifty-five new cases were registered under ICSID rules, a figure close to the previous year and among the highest in its history. The Centre also administered several hundred cases in total across the year, one of the largest annual #caseload figures it has ever recorded. These numbers confirm the description of the Convention as the backbone of the field. 9.2 What the disputes involve The pattern of cases tells its own story. ICSID's data show that energy and natural resources dominate. The oil, gas, and mining industries together with electric power and other energy sources account for a large share of new disputes, a reflection of the huge sums invested in these sectors and the political sensitivity that surrounds them. Infrastructure sectors such as transport and construction also feature heavily. The rise of energy cases in particular is tied to the global shift in energy policy, as governments change subsidy schemes and licensing rules and investors argue that those changes broke earlier promises. The geography is broad. Recent statistics show disputes involving states across Eastern Europe and Central Asia, South America, Central America and the Caribbean, Sub-Saharan Africa, and Western Europe, among others. This spread underlines an important point about the modern system: it is no longer a matter of wealthy investors suing only poor states. Developed economies now appear regularly as respondents, and the European Union itself became a party to an ICSID proceeding for the first time in the mid-2020s. 9.3 Who wins A common assumption is that the system is rigged in favour of investors. The data present a more balanced picture. ICSID's own figures indicate that outcomes are fairly even over the long run. Among cases decided by tribunals in recent years, roughly half of the awards upheld the investor's claim in whole or in part, while the rest either rejected the claim on the merits or found that the tribunal had no jurisdiction. A meaningful share of concluded cases are also settled or discontinued before any final decision. These figures do not settle the political argument about whether the system is fair, but they undercut the simplest version of the claim that investors always win. There is, however, an important qualification hidden inside these balanced numbers. Winning a case and being able to fight it are not the same thing. A state with a large legal budget and experienced government lawyers is in a very different position from a small economy that must hire foreign counsel at great expense and cannot easily match the resources of a well-funded claimant. The concern of #developing_countries is therefore less about the headline win rate and more about the cost of participation and the risk of a single very large award falling on a limited public purse. This is precisely why the proposal for an Advisory Centre, discussed later, has attracted so much support: it aims to level the field not by changing who wins, but by making it possible for poorer respondents to defend themselves properly in the first place. For students, the lesson is that fairness in this system has two layers. There is fairness in the outcome, which the data suggest is reasonably balanced, and there is fairness in access, which remains genuinely uneven. A complete assessment must weigh both. Criticisms and the Legitimacy Debate The very success of the Convention has drawn intense scrutiny. Over the past fifteen years a broad discussion, often called the #legitimacy_crisis or legitimacy debate, has questioned whether investor-state arbitration is a suitable way to resolve disputes that touch on public policy. The criticisms are serious and deserve careful statement, because reform cannot be understood without them. 10.1 Inconsistency and the lack of an appeal The first criticism follows directly from the design described earlier. Because there is no appeal and no doctrine of binding precedent, tribunals can reach conflicting conclusions on identical legal questions. Investors and states alike complain that they cannot predict how a standard treaty term will be read. Some scholars regard this inconsistency as the single most damaging feature of the system, because a body of law that cannot be predicted struggles to command respect. 10.2 Transparency A second criticism concerns #transparency. For much of its history investor-state arbitration was largely private. Hearings were closed, documents were confidential, and even the existence of a case might not be public. Critics argued that this secrecy was inappropriate for disputes in which the decisions could affect taxpayers, the environment, and public services. If a tribunal orders a government to pay a sum equal to a large part of its health budget, the public arguably has a right to know how and why. 10.3 The regulatory chill A third and politically charged criticism is the idea of the #regulatory_chill. The concern is that the mere risk of a large claim may discourage governments from adopting legitimate measures to protect public health, the environment, or financial stability. If a state fears that a new tobacco law, a mining ban, or a climate policy could trigger a claim for many millions of dollars, it may hesitate to act. Whether this chilling effect is real and how strong it is remain contested, but the perception itself has damaged confidence in the system among civil society groups. 10.4 Asymmetry and the public interest A fourth criticism is structural asymmetry. Under most treaties only the investor can bring a claim; the state can only defend. The system protects the investor's rights but says little about the investor's duties, such as obligations toward the environment or local communities. Critics argue that a mechanism concerned with the #public_interest should allow states and affected populations a fairer place in the process, rather than treating the dispute as a private matter between a company and a government. 10.5 Arbitrator independence and double-hatting A fifth criticism focuses on the arbitrators themselves. Because tribunals are assembled case by case from a relatively small professional community, the same individuals often appear repeatedly. A particular concern is #double_hatting, where a person acts as an arbitrator in one case while serving as counsel arguing a related point in another. Critics say this blurs the line between judge and advocate and threatens #arbitrator_independence. Others respond that the pool of genuinely expert individuals is naturally limited and that clear disclosure rules can manage the risk. 10.6 Cost and delay A final and very practical criticism is that arbitration has become slow and expensive. Cases can run for years and generate legal costs in the millions. This burden falls hardest on smaller enterprises and on states with limited budgets, and it can put the system out of reach for exactly the parties it might most help. Cost and delay are less dramatic than the constitutional concerns, but they may matter more to the everyday user. 10.7 Denunciation as protest These criticisms are not merely academic. Some states have acted on them. A small number of governments in Latin America, including Bolivia, Ecuador, and Venezuela, formally withdrew from the Convention, an act known as #denunciation, arguing that the system infringed their sovereignty. It is telling, however, that at least one of them, Ecuador, later rejoined, which suggests that leaving the system carries costs of its own by making a country less attractive to investors. The pattern of exit and return captures the dilemma at the centre of the whole debate: states resent the constraints of the system but fear the consequences of abandoning it. 10.8 The European dimension A distinct challenge came from inside Europe. The Court of Justice of the European Union held, in a decision widely known by the name #Achmea, that arbitration clauses in investment treaties between EU member states were incompatible with EU law. Most member states then agreed to terminate those internal treaties. This did not affect ICSID's role in disputes involving parties outside the Union, but it created deep uncertainty about a whole category of intra-European cases and added a further layer to the reform discussion. The Reform Programme Faced with these pressures, the system did not stand still. Since the late 2010s a wide and serious reform programme has been underway, operating at several levels at once. It is useful to divide it into changes ICSID made to its own rules and the broader, more ambitious work taking place through the United Nations. 11.1 The 2022 procedural rules ICSID undertook the most extensive revision of its procedures in its history, and the new #2022_Arbitration_Rules came into force on 1 July 2022 after a consultation process that lasted about five years and involved member states and many other participants. The old set of rules grew from around fifty-six provisions to roughly eighty-six, and the changes targeted several of the criticisms described above. The reforms attacked cost and delay directly. Tribunals and parties are now placed under an express duty to conduct proceedings in good faith and in a cost-effective and expeditious manner, with firm deadlines for key steps and mandatory case management conferences. A new option of #expedited_arbitration allows willing parties to run a streamlined procedure that can, in principle, cut the length of a case substantially, which is especially useful for smaller disputes. The reforms also improved transparency. The rules now provide a clearer path to publishing awards and decisions. Although publication of an award still depends on party consent, the rules introduce a form of deemed consent: if a party does not object within sixty days, consent is presumed, and where consent is withheld the Centre may still publish anonymised excerpts. Procedural orders and hearings are also treated with a stronger presumption of openness. For the first time the rules squarely address #third_party_funding, requiring a party to disclose whether a non-party is financing the claim and the identity of that funder, which responds to concerns that hidden backers could distort the process. The rules also clarified the procedure for security for costs and for the swift dismissal of claims that are manifestly without legal merit. Alongside the arbitration reforms, ICSID introduced dedicated rules for #mediation and for fact-finding. These give parties a structured way to seek a negotiated #settlement rather than fighting a full arbitration, reflecting a growing recognition that not every investment dispute needs to end in a binding award. The widening of the Additional Facility, discussed earlier, was part of the same package. 11.2 The UNCITRAL process The most far-reaching reform work is taking place not at ICSID but at the United Nations, through the #Working_Group_III of the United Nations Commission on International Trade Law, usually called #UNCITRAL. Since receiving its mandate in 2017, this group has brought governments together to consider whether and how to reform investor-state dispute settlement as a whole. Its work is deliberately open to every state, not only the users of any one institution, which gives it a legitimacy that a single body cannot claim. The Working Group organised the concerns into three broad clusters: the cost and duration of proceedings, the consistency and correctness of decisions, and the independence and diversity of the decision-makers. Around these clusters it has developed a menu of possible reforms, ranging from modest procedural adjustments to fundamental structural change. Several concrete outputs have already emerged. Working with ICSID, the group produced a #Code_of_Conduct for arbitrators and judges in investment disputes, adopted in 2023, which sets standards on independence, impartiality, disclosure, the limitation of double-hatting, and reasonable fees. The group has also advanced the idea of an #Advisory_Centre to help developing countries and smaller states handle the legal complexity and expense of these disputes, addressing the imbalance of resources that many respondents face. 11.3 Structural reform: appeals and a standing court The boldest proposals under discussion would change the shape of the system itself. Two ideas stand out. The first is an #appellate_mechanism, a body that could review awards for legal error and so bring greater consistency to the case law, directly answering the criticism that flows from the absence of an appeal. The second, more ambitious idea, championed in particular by the European Union, is a #multilateral_investment_court: a permanent tribunal with tenured members who would decide investment disputes in place of ad hoc tribunals. Supporters argue that standing judges, appointed by states rather than by the parties to each case, would be more independent and would produce a more coherent and legitimate body of law. Opponents worry that such a court could be slower, more politicised, or less expert than the current system, and that it would sacrifice the party control that makes arbitration attractive. These proposals remain under negotiation, and their fate will shape the field for a generation. 11.4 Two philosophies of reform Underlying the whole programme is a genuine disagreement about how far change should go. One camp favours incremental reform: keep the existing structure and fix its flaws through better rules, codes of conduct, and perhaps an appeal. Another camp favours systemic or even paradigmatic reform: replace ad hoc arbitration with a standing court, or rethink the balance between investor rights and state duties from the ground up. A third view resists reform altogether, arguing that the system already delivers balanced results and that dramatic change would create new problems. The reform debate is therefore not a single project but a contest between competing visions of what international investment law should be. Assessing ICSID as the Investor's Forum 12.1 Why the label fits, and why it is incomplete The Convention is often called the investor's forum, and the label captures a real truth. Before 1965 an investor abroad was legally weak, dependent on unreliable diplomatic protection and suspicious local courts. After 1965, and especially after the treaty network of the 1990s, an investor could enforce its rights against a sovereign state before a neutral tribunal and collect on the result in courts around the world. That is a genuine shift of power, and it is why the system is prized by those who invest across borders. Yet the label is incomplete in two ways. First, as the caseload data show, states win a substantial share of cases, so the forum is not simply a machine for transferring money to companies. Second, the system was never designed to serve investors alone. Its deeper purpose was to encourage #cross_border_investment for the sake of #economic_development, on the theory that capital will flow more freely to countries that offer a credible promise of fair treatment. On this view the ultimate beneficiary is meant to be the host state and its population, who gain the jobs, infrastructure, and technology that investment can bring. Whether the system actually delivers that development benefit is one of the hardest and least settled questions in the field. 12.2 The achievement and the tension The lasting achievement of the Convention is that it solved a problem of trust. It built a neutral space where a company and a government could resolve a dispute without either side feeling that the deck was stacked. It did this not by imposing rules on states but by offering a procedure they could accept, and by anchoring that procedure in the credibility of the World Bank. The breadth of participation and the size of the modern caseload are the measure of that success. The tension at the heart of the system is that the features which made it acceptable to states in 1965 are the same features now under attack. Confidentiality made governments comfortable but now looks like secrecy. Party-appointed arbitrators gave both sides control but now raise doubts about independence. Finality without appeal delivered certainty but produced inconsistency. The reform programme is, in essence, an attempt to preserve the neutrality and enforceability that give ICSID its value while removing the opacity and unpredictability that threaten its legitimacy. The difficulty is that these goals can pull in opposite directions, and the balance struck over the next decade will determine whether the framework created in Washington endures in its present form or gives way to something new. Conclusion The #Washington_Convention of 1965 is a short treaty with a long reach. It did not try to tell states how to treat foreign investors. It did something more subtle and, in the end, more powerful: it created a trusted process and let the substance grow up around it. Through the thousands of investment treaties that adopted ICSID as their forum, the Convention became the backbone of #investor_state_arbitration and the reason a private company can today hold a sovereign state to account before a neutral tribunal. For a student meeting the subject for the first time, the framework can be reduced to a few connected ideas. Everything rests on #consent, which is usually assembled from a treaty promise and a later acceptance. #Jurisdiction under Article 25 turns on the identity of the parties and on the contested meaning of #investment. The tribunal applies both national and international law under #Article_42 and produces a binding #award. That award cannot be appealed but can be challenged through the narrow route of #annulment under Article 52, and it is enforced across the membership under #Article_54, subject only to the surviving rule of #sovereign_immunity from execution. Around this core sit the Additional Facility, a large and balanced caseload, a sharp debate about #legitimacy, and a reform programme running from the #2022_Arbitration_Rules to the ambitious work of #UNCITRAL and the proposals for an #appellate_mechanism and a standing court. The Convention has survived sixty years because it answered a real need with a design of unusual restraint. Its future will depend on whether the same restraint can be shown by the reformers who now hold its fate in their hands. If they can keep what works, the neutrality and the enforceable promise, while curing what does not, the secrecy and the inconsistency, then the investor's forum built in Washington will remain the central institution of international investment law for many years to come. If they cannot, the framework of 1965 may prove to be the foundation on which a very different structure is built. #ICSID #Washington_Convention_1965 #investor_state_arbitration #international_investment_law #ISDS_reform #foreign_direct_investment #World_Bank #arbitration #enforcement_of_awards #Article_25_jurisdiction #UNCITRAL_Working_Group_III #multilateral_investment_court #bilateral_investment_treaties #dispute_settlement #depoliticization_of_disputes References Alschner, W. (2022). Investment Arbitration and State-Driven Reform: New Treaties, Old Outcomes. Oxford: Oxford University Press. Damjanovic, I. (2023). The European Union and International Investment Law Reform: Between Aspirations and Reality. Cambridge: Cambridge University Press. Dolzer, R., Kriebaum, U., and Schreuer, C. (2022). Principles of International Investment Law (3rd ed.). Oxford: Oxford University Press. https://doi.org/10.1093/law/9780192857804.001.0001 Franck, S. D. (2023). Reforming World Bank Dispute Resolution: ICSID in Context. Michigan Journal of International Law, 44(3), 349-406. Roberts, A., and St John, T. (2022). Complex Designers and Emergent Design: Reforming the Investment Treaty System. American Journal of International Law, 116(1), 96-149. Schill, S. W., Malintoppi, L., Reinisch, A., Schreuer, C. H., and Sinclair, A. (eds.) (2022). Schreuer's Commentary on the ICSID Convention: A Commentary on the Convention on the Settlement of Investment Disputes between States and Nationals of Other States (3rd ed.). Cambridge: Cambridge University Press. https://doi.org/10.1017/9781316516584

The #Riyadh_Convention of 1983, formally the Riyadh Arab Agreement for Judicial Cooperation, is the most comprehensive regional treaty binding the courts and prosecutors of the Arab world together. Adopted under the League of Arab States, it created shared rules for four connected tasks: sharing #legal_information, serving court papers and running #letters_rogatory across borders, recognizing and enforcing foreign #judgments and #arbitral_awards, and handling the #extradition of accused and convicted persons. This article gives students a clear, evidence based reading of the treaty. It explains how the agreement is built, what each part is meant to do, and how national courts and security bodies apply it in practice. It then weighs the treaty's strengths against its weaknesses. On the positive side, the Convention removes many old obstacles to #cross_border_justice, lowers the cost of litigation, and gives businesses and families more legal certainty across the region. On the negative side, weak drafting, an open ended #public_order clause, a narrow #political_offence_exception, and the absence of a clear #non_refoulement rule leave room for abuse. The paper pays special attention to the treaty's link with the Arab Interior Ministers Council and shows how the same machinery that helps ordinary justice can also serve #transnational_repression. The study concludes that the Convention is a useful but unfinished instrument that needs a shared reading, stronger #human_rights safeguards, and a proper oversight body. Keywords: Riyadh Convention 1983; judicial cooperation; extradition; enforcement of judgments; mutual legal assistance; Arab League; human rights; transnational repression 1. Introduction Justice does not stop at a national border, but the power of a court usually does. A judgment written in one country has no automatic force in another. A witness who lives abroad cannot be forced to appear. A suspect who crosses a frontier may slip out of reach. For a long time states solved these problems one deal at a time through bilateral treaties, which is slow and uneven. Regional agreements try to fix the whole map at once by giving many states a single shared rulebook. The #Riyadh_Convention is the Arab world's main attempt to do exactly this. It was signed in Riyadh, the capital of #Saudi_Arabia, on 6 April 1983, and endorsed by the Council of Arab Ministers of Justice under the umbrella of the #Arab_League. It came into force in 1985 and has since been ratified by most Arab states. The treaty covers a wide field: the exchange of #legal_information, service of judicial and non judicial documents, requests for evidence known as #letters_rogatory, the attendance of witnesses and experts, the #recognition_of_judgments and their enforcement, the enforcement of #arbitral_awards, the enforcement of criminal sentences, and the #extradition of persons wanted for prosecution or already convicted. The subtitle of this article, "the judicial policeman," captures a tension that runs through the whole instrument. On one reading the Convention is a service provider. It helps courts talk to one another and helps ordinary people and companies get results that would otherwise be blocked by distance and formality. On another reading the same treaty acts like a regional police force. It moves arrest requests, wanted lists, and human beings across frontiers, and it does so through security ministries as much as through courts. Whether the treaty is mainly a tool of cooperation or mainly a tool of control depends on how its clauses are drafted, how they are read by national judges, and what safeguards surround them. This article is guided by three questions. First, what does the treaty actually say, and how is it organized, so that a student can understand the mechanics without a specialist background? Second, how well does each mechanism work, and where does the drafting create gaps? Third, what does the record show about how the treaty and its related machinery have been used, including the documented cases in which people were sent back to states where they faced serious harm, and what reforms would keep the useful parts while closing the dangerous ones? The argument developed here is straightforward. The #Riyadh_Convention is a genuine achievement of regional #legal_integration and it solves real problems. Yet it was written in a way that leaves too much to national discretion and says too little about the rights of the individual. As a result it can serve justice and it can serve repression with the same set of tools. The fix is not to abandon the treaty but to modernize it, by adding precision, safeguards, and oversight to a structure that is otherwise sound. 2. Historical and Institutional Background Arab legal cooperation did not begin in 1983. The League of Arab States, created in 1945, produced several earlier instruments, including the 1952 Arab League Convention on the Enforcement of Judgments and an earlier agreement on extradition. These first efforts were narrow and were not widely applied. States kept relying on the older habit of bilateral treaties, each with its own conditions, which produced a patchwork that was hard for courts to use with confidence and that often broke down when two states had no treaty at all. By the late 1970s two pressures pushed the Arab states toward a broader deal. The first was the growth of cross border trade, labor migration, and family life spread across several countries, all of which produced disputes that no single national court could resolve on its own. A worker employed in one Gulf state, with a family in another and property in a third, generates legal questions that only a shared framework can answer smoothly. The second pressure was the rise of cross border crime and the movement of suspects between neighboring states, which made a common approach to #extradition and criminal assistance attractive to governments. The Council of Arab Ministers of Justice, a specialized body of the #Arab_League, took up the task of drafting a single treaty that would replace the older, fragmented arrangements. The result was the #Riyadh_Convention, designed to supersede the 1952 judgments convention in relations between states party to both, a point the treaty itself makes near its end. The number of parties is usually given as around eighteen to twenty one Arab states, and the exact count is disputed because signing and ratifying are not the same thing. Commonly listed parties include Jordan, Bahrain, Tunisia, Algeria, Djibouti, #Saudi_Arabia, Sudan, Syria, Somalia, Iraq, Oman, Palestine, Qatar, Kuwait, Lebanon, Libya, Morocco, Mauritania, Yemen, and the United Arab Emirates. Some states signed but never ratified, so the treaty does not bind them. Kuwait, for example, is often described as a signatory that did not complete ratification, which has caused confusion in national courts that assumed the treaty applied and cited it in cases where it did not govern. The position of a few other states is also debated in the scholarship. The practical lesson is that a court must check the exact status of each state, both signature and ratification, before applying the Convention, because getting this wrong can invalidate an entire decision. Two institutional features matter for the rest of this article. First, the Convention channels many of its procedures through Ministries of Justice, which allows courts and prosecutors to talk directly to one another rather than through slow diplomatic notes. This direct route is one of the treaty's real advances, because it cuts the delay that had made earlier cooperation almost useless in urgent matters. Second, on the criminal side the Convention connects to a security institution. It requires parties to coordinate #extradition procedures with what it called the Arab Organization for Social Defense Against Crime, also known as the Arab Criminal Police Bureau. That bureau's functions were later absorbed by the General Secretariat of the Arab Interior Ministers Council, a body founded in 1982 that groups the interior ministries of the member states and circulates arrest warrants on their behalf. This is the point where the treaty stops looking only like a service between courts and starts looking like a regional #law_enforcement network, a theme this article returns to in Section 9. 3. Method and Scope This study is a doctrinal and evaluative analysis. Doctrinal means that it reads the text of the treaty closely and works out what each clause requires, permits, or forbids. Evaluative means that it does not stop at description but judges how well the rules function, using three tests: does the mechanism achieve its stated aim, is it clear enough to be applied consistently, and does it protect the rights of the people it affects. The evidence base has three layers. The first is the treaty text itself, in the widely used unofficial English translation, which is the primary source for every claim about what the Convention says. The second is secondary legal literature and professional commentary on how national courts in the region apply the treaty, especially on the #enforcement_of_judgments and on service of process, including reported decisions from the Gulf. The third layer is the record kept by human rights organizations, national courts, and inter governmental bodies on how the criminal cooperation machinery has operated in real cases. A word of caution belongs here. The treaty exists in Arabic as the authentic text, and translations differ in small but sometimes important ways, including the numbering and wording of articles. Where this article refers to a specific article number, it does so as a guide rather than as a precise citation, because published sources do not always agree on numbering. The safest approach for any practitioner is to consult the authentic Arabic text together with the national instrument of ratification, since the status and even the wording that binds a given state can vary. 4. Structure and Scope of the Agreement The #Riyadh_Convention is best understood as a bundle of separate regimes held together in one document. Each regime handles a different stage in the life of a legal problem that crosses a border. Reading the treaty this way makes it far easier to follow than reading it article by article. The first regime is the exchange of #legal_information. States agree to send one another the texts of their laws, information about their court systems, and other material that helps a foreign court understand and apply the law of a partner state. This is the quiet foundation on which everything else rests, because a court cannot respect a foreign judgment or a foreign law it does not understand. The second regime covers the #service_of_documents, both judicial and non judicial. When a lawsuit in one country must reach a defendant living in another, the papers have to be delivered in a way the law recognizes. The Convention creates a direct method, often called Convention Service, in which a court in the sending state passes documents to a court in the receiving state, cutting out slow diplomatic channels. This method has been examined closely by courts in the region, including in disputes about whether it is the only permitted route or merely one option among several. The third regime is the #letters_rogatory system, also called rogatory commissions. This lets a court in one state ask a court in another to take a step on its behalf, such as hearing a witness, collecting a document, or carrying out an inspection. The requested court runs the task under its own procedural law, and the requesting court gets usable evidence without having to send its own officials abroad. The fourth regime governs the attendance of witnesses and experts, especially in criminal matters, with protections so that a witness who travels to help a foreign court is not arrested or prosecuted for earlier matters while doing so. This safe conduct rule encourages people to cooperate with justice across borders, which they would not do if the trip itself put them at risk. The fifth regime, and one of the most used in commercial life, is the #recognition_of_judgments and their enforcement, together with the enforcement of #arbitral_awards. Here the treaty tells a court in one state how and when it must give effect to a decision made by a court or arbitral tribunal in a partner state. The sixth regime deals with criminal judgments, allowing final criminal sentences passed in one state to be enforced against that state's own nationals in another, under conditions such as a minimum remaining sentence of six months and the requirement that the offence be one for which extradition is available. The seventh and most sensitive regime is extradition, the surrender of a person wanted for trial or already convicted, from the state where the person is found to the state that wants them. Two design choices shape all of these regimes. The first is the principle that the requested court does not re examine the merits of the foreign decision. Its job is to check that the conditions of the treaty are met, not to retry the case. This keeps the system efficient and respects the work of the first court. The second is a set of escape hatches, most importantly a #public_order clause and, on the civil side, a reference to the rules of Islamic #Sharia, which allow a requested state to refuse cooperation that would offend its fundamental legal values. These escape hatches are necessary in any treaty of this kind, but their width is where much of the trouble begins, as later sections show. 5. Exchange of Legal Information The exchange of legal information is the least dramatic part of the treaty and the easiest to overlook, yet it does important work. Under this regime the parties commit to supply one another, on request, with the texts of laws in force, information about judicial organization, and other legal material needed to apply or understand the law of a partner state. The requests move through the Ministries of Justice, which keeps the channel official and reliable. Why does this matter in practice? Consider a judge in one Arab state asked to enforce a contract governed by the law of another. To do the job properly the judge may need to know the exact wording of a limitation period, or whether a particular type of clause is valid under the other state's law. Without a dependable way to obtain that information, the judge is left to guess, and guessing produces inconsistent and unfair results. A working information channel reduces that risk and supports every other part of the Convention, because the #recognition_of_judgments, the letters rogatory process, and even extradition all assume that the requested authority can find out what the requesting state's law actually says. The main weakness of this regime is that it depends on administrative goodwill and capacity. A commitment to share information is only as good as the office that answers the request. Where a Ministry of Justice is under resourced or slow, the channel becomes a bottleneck. There is also no strong deadline or enforcement mechanism, so a state that drags its feet faces no real consequence. In an age of digital legal databases, much of this could be solved by a shared regional portal of laws and court structures, which would turn a slow paper process into an instant lookup. The treaty predates that technology, and updating this regime would be a low cost, high value reform that touches none of the sensitive parts of the treaty. 6. Service of Documents and Letters Rogatory For a lawsuit to be fair, the person being sued must be told about it in good time and in a proper way. When the defendant lives in another country, this simple requirement becomes a real obstacle. The Riyadh Convention tackles it through its rules on the #service_of_documents. The treaty allows documents in civil and commercial matters to move between the competent authorities of the two states, and it treats service carried out this way as if it had happened in the requesting state itself. In criminal matters the papers travel directly between Ministries of Justice, with copies to the foreign ministries, which keeps a diplomatic record without slowing the process to a crawl. The direct, court to court method has been the subject of important litigation, because a lot turns on whether it is the exclusive route or simply the preferred one. In one closely studied line of cases in the Gulf, the courts of the Dubai International Financial Centre had to decide whether a claimant who could not complete Convention Service could instead ask for an alternative form of service, or whether the treaty method was mandatory and could not be bypassed. An earlier decision had suggested the treaty method was exclusive, which created serious practical problems, because a foreign local court could effectively block a case simply by refusing to serve the papers. On appeal the court held that the treaty method is a real and useful channel but does not shut out other lawful ways of giving notice, and it clarified that obligations in international agreements on civil and commercial matters are not automatically part of the free zone's own law. That gave courts needed flexibility when a foreign authority is slow or unhelpful. The lesson for students is that the value of a service rule depends not only on its wording but on how courts read the word "may" against the word "must." The second mechanism here is the letters rogatory system. A rogatory commission is a formal request from a court in one state to a court in another to perform a specific judicial act, most often the taking of evidence. The Convention sets out what the request must contain, including the type of action sought, the names of the requesting and requested bodies, and the facts of the case, along with the names of witnesses and the questions to be put to them. Importantly, the treaty relieves the parties of heavy certification requirements, so the request and its enclosures do not need to be formally legalized, which cuts delay and cost. The requested court carries out the task under its own procedural law and is generally obliged to comply, with only narrow grounds for refusal, such as where the request would harm the sovereignty or security of the requested state. Taken together, these two mechanisms are among the most practical and least controversial parts of the treaty. They rarely raise the deep #human_rights problems that extradition raises, and they deliver a steady stream of ordinary benefits: enforceable notice, usable evidence, and lower litigation costs. Their main limits are administrative rather than legal. Different states apply different standards of speed and diligence, and there is no central body to monitor performance or resolve disputes about how a request was handled. A modest reform, such as model forms, agreed service standards, and a shared electronic transmission system, would improve reliability without touching the sensitive parts of the treaty. 7. Recognition and Enforcement of Judgments and Arbitral Awards For businesses, families, and lawyers, the #enforcement_of_judgments is the part of the Riyadh Convention that matters most in daily life. A judgment that cannot be enforced where the losing party keeps their assets is close to worthless. This regime, covering roughly the middle articles of the treaty, tells a court in one Arab state when and how it must give effect to a judgment from a partner state, and it extends the same logic to #arbitral_awards. The core rule is efficient and pro enforcement. When a party asks a court to recognize or enforce a foreign judgment, the court's duty is limited to checking that the judgment meets the treaty's conditions. It must not reopen the dispute or examine the merits. In other words, the requested court is a gatekeeper, not a court of appeal. Once the conditions are satisfied, the court gives the foreign judgment the same enforceable status it would have had if a local court had made it, and it can even enforce part of a judgment if that part can stand on its own. This approach lines up with the modern trend in #private_international_law toward the free movement of judgments, and scholars who study the region regard the treaty as the most significant Arab instrument in this field, even while noting that it has received surprisingly little detailed academic attention. The treaty also lists the grounds on which recognition may be refused, and these grounds define the real shape of the regime. A judgment may be refused if it conflicts with Islamic #Sharia, the constitution, #public_order, or the basic rules of conduct of the requested state. It may be refused if it was given by default without proper notice to the absent party, so that the person had no fair chance to defend the case. It may be refused if the rules on legal representation of persons who lack full legal capacity were ignored, and it may be refused if the same dispute has already produced a final judgment in the requested state, or in a third state that the requested state recognizes. Certain categories, such as provisional and precautionary measures and matters like bankruptcy, taxes, and fees, sit outside the ordinary enforcement channel. Each of these refusal grounds is reasonable on its own. Together, though, they create the treaty's central weakness on the civil side: the width of the public order and #Sharia exceptions. Because these terms are not defined and are read through the eyes of the requested state, a court that does not want to enforce a particular judgment has an easy tool to refuse it. This turns a system that looks automatic on paper into one that can be unpredictable in practice. The remedy, as commentators repeatedly stress, is a shared or harmonized reading of these clauses, ideally guided by a regional body, so that public order means roughly the same thing from one capital to the next. Without that, the promise of near automatic enforcement is only partly kept. A second, quieter problem is the requirement, in several states, of #reciprocity, together with confusion in national courts about the difference between recognition and enforcement. Recognition means treating a foreign judgment as settling the matter between the parties. Enforcement means using the state's coercive power to carry the judgment out, for example by seizing assets. The two ideas call for different procedures, and courts in the region have sometimes mixed them up, which produces avoidable delay and error. Recent decisions in the Gulf show national courts slowly clarifying these concepts and, in some rulings, moving toward a more liberal and predictable approach to indirect jurisdiction and enforcement, although the movement is uneven and some courts have swung back to more restrictive positions. That back and forth shows that the treaty text alone does not settle the matter; consistent judicial practice has to grow around it. The enforcement of arbitral awards deserves separate mention because of its economic weight. Arbitration is the preferred way of resolving large commercial disputes in the region, and an award is only valuable if it can be enforced where the assets are. The Convention allows arbitral awards made in one party state to be recognized and enforced in another, subject to the same style of limited review and the same public order and Sharia safeguards. This complements, rather than replaces, the global New York Convention on the recognition of foreign arbitral awards, to which many Arab states also belong. For a business choosing where to arbitrate and how to structure a deal, the combined effect of these instruments is greater legal certainty and, therefore, greater confidence to invest across borders. That link between enforceable judgments and economic integration is one of the strongest arguments in the treaty's favor. There is also an important question of how the treaty sits inside modern financial free zones such as the Dubai International Financial Centre and the Abu Dhabi Global Market, which run on common law style rules. Because these zones apply their own procedural law, courts have had to work out when a national treaty like the Riyadh Convention binds them and when it does not, and how a judgment from one of these zones can travel to another Arab state under the treaty. The answers are still developing, but the pattern is that treaty obligations forming part of the wider national law are respected within the zones on enforcement questions, even though not every treaty rule is automatically imported into the zone's internal procedure. For students, this is a useful example of how a single regional treaty interacts with unusual domestic structures. The service of process point deserves a final mention here because it links Sections 6 and 7. In several reported enforcement cases, the reason a foreign judgment failed was not that it was wrong on the merits but that the defendant had not been properly served. This is a reminder that the different regimes of the treaty are connected. A weak link in service can defeat an otherwise sound claim for enforcement, which is why improving the practical operation of service would also strengthen the enforcement regime. 8. Extradition #Extradition is the surrender of a person from the state where they are found to a state that wants them for trial or to serve a sentence. It is the sharpest exercise of cooperation, because it puts a human being, not just a document or a sum of money, into the hands of a foreign authority. The Riyadh Convention devotes a substantial block of articles to it and sets out both the duty to extradite and the exceptions to that duty. The general obligation is that parties will surrender persons found on their territory who are wanted by another party for prosecution or for the enforcement of a sentence, provided the case falls within the treaty and does not hit one of the exceptions. Around this core the treaty builds a set of familiar safeguards drawn from the general law of extradition. It applies a form of the #dual_criminality principle, meaning the conduct must amount to an offence in both states and usually must carry a sentence above a set threshold. It allows a state to refuse to surrender its own nationals, a common feature in civil law systems, which is why many Arab states will prosecute a national at home rather than hand them over. It regulates the situation where the wanted person is already being prosecuted or serving a sentence in the requested state, allowing surrender to be postponed until local proceedings end, while permitting a temporary transfer so the person can appear before the requesting court and then be returned. It contains a version of the rule of specialty, so that if the classification of the offence changes after surrender, the person may not be charged or tried for the new offence unless that offence would itself have justified extradition. And it coordinates the timing and paperwork of requests with the regional police machinery. The most important exception, and the most contested, is the #political_offence_exception. The treaty allows a requested state to refuse surrender where the offence is regarded under its own law as political in nature. This rule exists in almost every extradition treaty in the world for a good reason: states should not become instruments for punishing peaceful opposition abroad. The problem in the Riyadh text is twofold. First, the judgment about whether an offence is political is left to the discretion of the requested state, with no shared standard and no independent check, so the same facts can be political in one capital and ordinary crime in another. Second, and more seriously, the treaty and the related regional instruments carve important conduct out of the political category. Acts such as attacks on heads of state, and offences described as offending kings and rulers, are treated as ordinary crimes for which surrender is available. This carve out sharply narrows the protection that the political offence exception is supposed to give, because in practice much peaceful dissent in the region is prosecuted precisely under laws that protect rulers and state institutions, or under broadly written anti terrorism statutes. There is a further, structural gap. Modern extradition law is expected to include a clear rule of #non_refoulement, the principle that no one may be sent to a place where they face a real risk of torture or other grave harm. This principle is binding on Arab states through the Convention against Torture and other instruments, and it is echoed in the Arab Charter on Human Rights. The Riyadh Convention, however, does not spell out a non refoulement rule of its own and does not mention the risk of torture as a ground for refusal. It also permits, in some readings, the enforcement of judgments and surrender in cases decided in absentia, subject to conditions, even though a decision reached without the accused present raises obvious fairness concerns. The absence of an explicit non refoulement clause means that the treaty's safeguards depend entirely on national law and on the good faith of the requested state, which is a thin protection when the requesting state has a poor human rights record. The Convention also connects to the enforcement of criminal sentences. A final penal verdict from one state may be enforced against the convicted person in the state of their nationality, subject to conditions such as a minimum remaining term and the exclusion of offences for which extradition would not apply. In principle this can be humane, because it lets a prisoner serve a sentence closer to home, but it depends on the fairness of the original trial. Where that trial was unfair, transferring the sentence simply exports an injustice. The result is a regime that works smoothly for ordinary crime, which is its main and legitimate purpose, but that offers weak resistance when a request is politically motivated and dressed up as ordinary crime. This is the seam along which the treaty's benefits and dangers separate, and it is examined further in Sections 9 and 11. 9. The Judicial Policeman: Institutional Enforcement Through the AIMC and Interpol The phrase "judicial policeman" is not only a metaphor. The Riyadh Convention is wired into a real regional #law_enforcement network, and understanding that wiring is essential to any honest assessment of the treaty. The Convention tells parties to coordinate their extradition procedures with a regional police body, originally the Arab Criminal Police Bureau. Over time the functions of that bureau were absorbed into the General Secretariat of the Arab Interior Ministers Council, a specialized council of the #Arab_League that brings together the interior ministries of the member states. Through this body, a member state can circulate an arrest warrant to the interior ministries of the other members, which then act to locate and detain the wanted person with a view to surrender. Liaison offices inside each ministry keep the traffic moving. In effect, the treaty's criminal side runs on a security track operated by interior ministries, in parallel with the judicial track operated by justice ministries and courts. This design has a serious consequence. Unlike #INTERPOL, the global police organization, the regional council does not appear to have an independent oversight body that filters out abusive requests, and it does not offer targeted individuals a clear way to see the warrant against them, challenge it, or have it removed. #INTERPOL, for all its own well documented problems, at least has a constitution that forbids activity of a political, military, religious, or racial character, a task force that screens notices before publication, and a commission that can review and delete improper records. In one recent year that task force refused well over a thousand notices and diffusions for non compliance, several hundred of them for breaching the rules against political abuse. The regional Arab council, as human rights researchers point out, lacks equivalent safeguards, which makes its warrants far harder to contest. The two systems also interact. The regional council and #INTERPOL cooperate under a memorandum of understanding, and a request that starts in one channel can gain force through the other. INTERPOL's own #Red_Notice system, a request to police worldwide to locate and provisionally arrest a person pending extradition, has been widely criticized for abuse by states seeking to reach dissidents abroad. Scholarship on transnational policing describes how some governments weaponize the Red Notice mechanism against political opponents, and how democratic states then face a hard choice between honoring police cooperation and refusing to become complicit in repression. When a politically motivated request travels through both the regional council and INTERPOL, the person targeted can be exposed to arrest across a very wide area, sometimes far outside the Arab region, as several documented cases show. The core point is not that regional police cooperation is bad. Coordinated action against genuine cross border crime is necessary and valuable, and it saves lives. The point is that the Convention built a powerful enforcement engine and attached it to security ministries, without building an equally strong braking system of independent review and individual remedy. An engine without brakes is efficient in the wrong direction as easily as the right one. This imbalance between capacity and control is the heart of the "judicial policeman" problem, and it is the single feature that most sharply distinguishes the Arab framework from better balanced systems elsewhere. 10. The Convention and National Legal Orders A regional treaty is only as strong as its reception into domestic law, and here the Arab states differ. Some legal systems treat a ratified treaty as directly part of national law that judges can apply at once, while others require a separate act of legislation before the treaty has internal force. The Riyadh Convention therefore does not operate the same way everywhere. In one state a litigant can rely on the treaty directly; in another the same litigant must point to the national law that gives it effect. This difference explains why the treaty's real world impact is patchy, and why the same clause can be decisive in one country and almost invisible in another. The relationship becomes more complex where a state contains special legal enclaves. The Gulf financial free zones already mentioned are the clearest example. Because they apply their own procedural rules, courts there have had to decide when the national treaty binds them, and the answers turn on how the free zone's founding law imports national obligations. The general position that has emerged is that the treaty's enforcement obligations are respected where they form part of the surrounding national law, but that not every treaty rule is automatically part of the zone's own procedure. This kind of question, obscure as it sounds, has large commercial consequences, because it determines whether a judgment can move between an Arab state and a free zone court and back again. The interaction of overlapping instruments adds a further layer. A single dispute may fall within the Riyadh Convention, an older 1952 convention, a Gulf Cooperation Council convention, and a bilateral treaty all at once. When these instruments say different things, a court must decide which one governs, and the treaties themselves give little guidance on priority. National courts have sometimes applied the wrong instrument, or applied one while ignoring another that also fit, producing decisions that are hard to reconcile. For the treaty to work as intended, courts need clear rules on the order of priority among conventions, and litigants need to plead the correct instrument from the start. 11. Human Rights Tensions and Documented Cases The abstract concerns in the earlier sections become concrete when set against real cases. Human rights organizations, courts, and inter governmental bodies have documented a pattern in which the criminal cooperation machinery connected to the Riyadh Convention, especially through the Arab Interior Ministers Council, has been used to reach peaceful critics and return them to states where they faced grave harm. Between 2022 and 2025, researchers documented several cases of individuals targeted through the council's warrant circulation system, involving requesting and requested states across the region. The people affected were, in the documented cases, peaceful dissidents, protesters, or members of religious minorities, and the destinations were states where the returned person risked torture and an unfair trial. A few examples show the pattern. In early 2023, a Saudi national from the country's Shia minority was arrested at an airport in Morocco on the basis of a warrant circulated through the regional council, and after national court proceedings he was extradited to #Saudi_Arabia despite warnings that he faced a real risk of torture, with his fate afterward reported as unknown. In May 2023, a prominent Emirati dissident who also held another nationality was extradited from Jordan to the United Arab Emirates, in what human rights groups described as a breach of the #non_refoulement principle, after having been convicted in absentia in an earlier mass trial widely condemned as unfair. In January 2025, an Egyptian and Turkish poet was arrested in Lebanon and then extradited, on a request circulated through the regional council that cited vaguely worded security charges connected to a social media post, drawing condemnation from human rights organizations who called it politically motivated and contrary to Lebanon's own law barring extradition for political offences. In a related pattern outside the Arab region, a Gulf dissident was returned from a European state on the basis of an INTERPOL #Red_Notice, in circumstances that a regional human rights court had tried to halt only days earlier. Several features recur across these cases. The charges are often based on broad anti terrorism or national security laws that define terrorism vaguely enough to capture peaceful dissent. The requests travel through security channels that lack independent review. The #political_offence_exception fails to protect the target, either because the requested state reads it narrowly or because the alleged conduct falls into one of the treaty's carve outs. And there is no explicit #non_refoulement rule in the Convention to force the requested court to weigh the risk of torture before it acts. The combination turns a treaty designed for ordinary justice into a channel for #transnational_repression, defined as state action across borders to silence or punish dissent expressed from abroad, a phenomenon that United Nations bodies and independent researchers have documented in detail in recent years. It is important to keep proportion. The great majority of requests under regional cooperation concern ordinary crime, and the treaty delivers real public safety benefits when it helps return fraudsters, traffickers, and violent offenders to face justice. The problem is not that cooperation exists but that the same machinery has too few safeguards to distinguish a genuine criminal request from an abusive political one. When the safeguards are weak, even a small share of abusive requests can ruin lives and can chill the free expression of an entire diaspora, because people who fear that a warrant may follow them anywhere will censor themselves. This is why the human rights critique is not an attack on cooperation as such but a demand that cooperation be made accountable, and it is a critique that draws on the wider literature on the abuse of police cooperation worldwide, not only in the Arab region. 12. The Riyadh Convention in Comparative Perspective The strengths and gaps of the Riyadh Convention come into focus when it is placed beside other instruments that do similar work. The closest comparison is the 1996 Gulf Cooperation Council Convention for the Execution of Judgments, Delegations and Judicial Notifications, which binds the six Gulf states. The GCC and Riyadh instruments overlap heavily, covering the recognition of judgments, rogatory assistance, and service of documents, and they follow a similar logic of limited review and defined exceptions, with several refusal grounds worded almost identically. The main difference is scope: the GCC convention is narrower in membership but often deeper in integration among a smaller, more homogeneous group, while the Riyadh treaty is broader in membership but looser in application across a more diverse set of legal systems. Where both apply, courts must work out which instrument governs, and the interaction of overlapping conventions is itself a recurring source of confusion in national case law. A second comparison is with the European system, especially the 1957 European Convention on Extradition and the family of European instruments on mutual assistance and the movement of judgments. The European model is instructive because it pairs cooperation with a strong, shared #human_rights framework and, crucially, an external court, the European Court of Human Rights, that can review a surrender before or after it happens. European law also developed a firm non refoulement bar, so that surrender to a real risk of torture or a flagrantly unfair trial can be stopped, and it has moved toward mutual recognition of criminal decisions among trusted partners. The Riyadh system has the cooperation half of this equation but not the review half. It has no regional court that an individual can turn to in order to block an abusive extradition, and it does not contain an explicit non refoulement clause. The comparison suggests that the missing piece in the Arab framework is not more cooperation but more accountable cooperation. A third comparison is with the global work of the Hague Conference on Private International Law, including the 2019 Hague Judgments Convention and the older Hague instruments on service and evidence. These treaties show how modern drafting handles the same tasks with more precise definitions, clearer refusal grounds, and mechanisms designed to be read consistently across very different legal cultures. Arab scholars who study the 2019 Hague Judgments Convention from the perspective of the region note that Arab states could either join such global instruments or modernize their own treaties along similar lines. Either path would help solve the interpretation problem that weakens the Riyadh regime, because a shared standard reduces the room for a requested court to hide behind an undefined public order clause. The comparative picture is therefore encouraging in one respect and sobering in another. It is encouraging because the Riyadh treaty is clearly part of a worldwide movement toward easier #cross_border_justice, and its basic architecture is sound and comparable to instruments used elsewhere. It is sobering because the instruments that work best pair openness with oversight, and it is precisely oversight that the Arab framework lacks. 13. Challenges, Gaps, and Directions for Reform Drawing the threads together, the Riyadh Convention faces a set of connected problems, most of which can be addressed without discarding the treaty. The first challenge is interpretive inconsistency. The undefined public order and #Sharia exceptions, and the muddle in some courts between recognition and enforcement, mean that the same treaty produces different outcomes in different capitals. The fix is a shared reading. A standing committee under the #Arab_League, or a regular meeting of senior judges, could issue guidance on the meaning of the key exceptions and on procedure, giving national courts a common reference point. Scholars who study the enforcement regime argue that unlocking the treaty's potential depends precisely on the parties reaching a joint interpretation of its provisions rather than each reading them in isolation. The second challenge is the weakness of individual safeguards on the criminal side. The narrow #political_offence_exception, the absence of an explicit non refoulement rule, and the treatment of some in absentia decisions leave the person facing extradition with thin protection. The reform here is to amend the treaty, or to adopt a protocol, that adds a clear non refoulement clause barring surrender where there is a real risk of torture or grave harm, that requires a fair opportunity to be heard before surrender, and that narrows or removes the carve outs which strip peaceful political conduct of protection. The third challenge is the lack of oversight over the enforcement machinery. The regional police channel operated through interior ministries has no independent body to screen requests and no clear remedy for a person wrongly listed. The reform is to build one: an independent review mechanism, modeled loosely on INTERPOL's own commission but stronger, with the power to refuse or delete abusive warrants and to hear complaints from affected individuals. Pairing the enforcement engine with a real braking system is the single most important structural change the framework needs. The fourth challenge is administrative and technical. The exchange of legal information, the #service_of_documents, and the letters rogatory process all depend on slow paper procedures and uneven national capacity. The reform is modern infrastructure: shared electronic transmission, model forms, agreed service and response times, and a regional database of laws and court structures. These changes are cheap, uncontroversial, and would improve the treaty's daily usefulness at once. A fifth, cross cutting challenge is the tangle of overlapping instruments, including the older 1952 convention, the GCC convention, and a web of bilateral treaties, which forces courts to work out which rules govern a given case. Clearer rules on the priority of conventions, written into an updated Riyadh text, would reduce error and delay and would make outcomes easier to predict. None of these reforms requires the Arab states to abandon regional cooperation. On the contrary, they would make cooperation more trustworthy, and a more trustworthy system is a stronger one. States are more willing to enforce one another's judgments and honor one another's requests when they trust that the shared rules are fair and consistently applied. Reform is therefore not a concession to outside critics but an investment in the treaty's own effectiveness and in the #rule_of_law across the region. 14. Conclusion The Riyadh Convention of 1983 is the backbone of #judicial_cooperation in the Arab world. Across four decades it has given courts a common way to share information, deliver documents, gather evidence, recognize and enforce #judgments and arbitral awards, and surrender wanted persons. For businesses and families whose lives cross borders, it removes obstacles that no single national court could remove alone, and it stands, as scholars agree, as the most significant regional instrument of its kind for the Arab states. Yet the treaty was written in a way that leaves too much to national discretion and says too little about the individual. Its wide, undefined exceptions make the #enforcement_of_judgments less predictable than the text promises. Its narrow political offence exception and its silence on non refoulement make its extradition regime dangerous when a request is politically motivated. And its connection to a regional security network with no independent oversight gives the treaty the character of a judicial policeman, one that can serve justice and serve #transnational_repression with the same set of tools. The honest verdict is that the Convention is a useful but unfinished instrument. Its architecture is sound and worth keeping. What it lacks is precision, safeguards, and oversight. A shared reading of its key clauses, an explicit human rights floor including a firm non refoulement rule, an independent review body for the enforcement machinery, and modern administrative infrastructure would keep everything valuable in the treaty while closing the gaps that allow abuse. The regional will to cooperate already exists, as the treaty itself proves. The task now is to make that cooperation not only efficient but also accountable, so that the judicial policeman serves the law rather than the powerful. #Riyadh_Agreement_1983 #Riyadh_Convention #judicial_cooperation #Arab_League #extradition #enforcement_of_judgments #mutual_legal_assistance #letters_rogatory #recognition_of_judgments #non_refoulement #political_offence_exception #transnational_repression #private_international_law #human_rights #cross_border_justice References Amnesty International. (2026). Global: Misuse of Interpol Red Notices to Target Dissidents a Grave Institutional Failure. London: Amnesty International. Bromund, T. R. (2025). How the Abuse of INTERPOL Contributes to Transnational Repression. Washington, DC: New Lines Institute for Strategy and Policy. Congressional Research Service. (2024). Transnational Crime Issues: INTERPOL (Report IF10766). Washington, DC: Library of Congress. Elbalti, B. (2017). Reciprocity and the recognition and enforcement of foreign judgments: A lot of bark but not much bite. Journal of Private International Law, 13(1), 184-215. Elbalti, B. (2023). Perspective from the Arab world. In M. Weller et al. (Eds.), The 2019 HCCH Judgments Convention: Cornerstones, Prospects, Outlook (pp. 181-210). Oxford: Hart Publishing. Freedom House. (2021). Out of Sight, Not Out of Reach: The Global Scale and Scope of Transnational Repression. Washington, DC: Freedom House. Human Rights Watch. (2026). Lebanon: A Year Since Poet Unlawfully Extradited to the UAE. New York: Human Rights Watch. League of Arab States, Council of Arab Ministers of Justice. (1983). Riyadh Arab Agreement for Judicial Cooperation. Cairo: League of Arab States (unofficial English translation). MENA Rights Group. (2025). Understanding the Arab Interior Ministers Council and Its Role in Transnational Repression. Geneva: MENA Rights Group. MENA Rights Group and Cedar Center for Legal Studies. (2025). Joint Stakeholder Submission to the UN Human Rights Council: Universal Periodic Review of Lebanon, Fourth Cycle. Geneva: MENA Rights Group. Oxford Pro Bono Publico. (2025). Transnational Repression of Human Rights Defenders: References, Definitions and Protections under International Human Rights Law. Oxford: University of Oxford, Faculty of Law. San, S. (2022). Transnational policing between national political regimes and human rights norms: The case of the Interpol Red Notice system. Criminology and Criminal Justice. DOI: 10.1177/13624806221105280 Shaw, M. N. (2021). International Law (9th ed.). Cambridge: Cambridge University Press.