International Chamber of Commerce
Dispute Resolution in a Geopolitical World
Sir Geoffrey Vos, Master of the Rolls
Wednesday 14 October 2026 at 2pm
The future of cross border dispute resolution
Introduction
1. Many thanks for that kind introduction.
2. The title of this conference is Dispute Resolution in a Geopolitical World. For such a discussion, there are certain basic parameters that must underpin the debate.
3. First, dispute resolution mechanisms at all levels are in the process of seismic change. Those changes apply as much to domestic dispute resolution as they do to international commercial dispute resolution, as they do to court-based dispute resolution, arbitration, mediation and consensual AI- enabled dispute resolution. I shall return to that new, but rapidly growing concept.
4. Secondly, there are now many more options than there once were for international commercial parties. There are, of course, still the old stalwarts, the domestic courts most commonly chosen, including the Business and Property Division here in London and a few domestic courts elsewhere in the US, the Far East and Australasia. But there are also new International Commercial Courts, some well-established and some just appearing. They include the SICC, the DIFC, the Netherlands Commercial Court and new ones such as the Hong Kong International Commercial Court and many more. There are also a variety of models for these courts that has a bearing on their utilisation, but I do not want to divert into that complex topic now.
5. Thirdly, international commercial arbitration is more extensive and diverse than ever. There are literally thousands of arbitral institutions if you include sector specific ones. The main ones are well known and include, for example the ICC, the LCIA, the ICDR, SIAC, the HKIAC, ICSID and WIPO.[1]
6. All these dispute resolution mechanisms are affected by the seismic changes I have mentioned. The organisers summarise them as including heightened trade tensions, economic fragmentation, sanctions regimes, technological competition, armed conflict, which, they say, are increasing both the volume and complexity of disputes whilst testing the resilience, neutrality, and enforceability of arbitration and ADR mechanisms, particularly where conflict disrupts contracts, supply chains, and the rule of law.
7. I am going to focus on two connected changes that will potentially be fundamental and long-lasting: technology and enforcement. I believe that many practitioners are in denial about the scale of the changes that new technologies will make to the future of dispute resolution. And I believe that, once those changes start to crystallise, questions of enforcement will arise more frequently and more acutely than they have in the recent past.
8. So, let me start by setting the scene, before mapping out some of the changes that technology will bring to dispute resolution, and exploring a little how that may affect the global order.
Setting the scene
9. The first thing to understand is that arbitration is not the same as litigation so far as technology is concerned. It is often mistakenly considered in the same breath.
10. Arbitration is an almost entirely consensual process, but court-based dispute resolution is not, even if the forum and applicable law are chosen by the parties. States provide court-based dispute resolution as a part of their constitutional settlement. For that reason, courts and judges cannot depart from the established mechanisms for that process without legislative authority – at least in a democracy. This is critical when one comes to consider the greatest disruptor of our age, namely machine-made decision-making.
11. Judges and courts cannot just abdicate responsibility for their decisions to Claude or ChatGPT without legislative or other constitutional authority. Such authority is being considered or provided in a number of jurisdictions, but all those jurisdictions underpin their introduction of AI into the judicial decision-making process with the foundational statement that judges will always have the last word. They all say that AI can help the judge, but must never replace the judge. That may or may not prove to be the reality, but it is the premise on which almost all jurisdictions are operating.
12. But be that as it may, English and Welsh judges could not, I think, abdicate their decision-making responsibility to AI until Parliament said that they could and until it was clear that what replaced them was an “independent and impartial tribunal established by law” within the meaning of article 6 of the ECHR.
13. Conversely, arbitrators can use machines to assist their decision-making in any way that the parties may agree. We know that the International Centre for Dispute Resolution (ICDR) of the American Arbitration Association (AAA) announced in December 2025 the availability of an AI arbitrator for two party, document-only construction disputes up to a certain value (perhaps $50,000).[2] Whilst the model includes a “human in the loop”, the AI arbitrator will evaluate the merits of claims and prepare draft awards. The human arbitrator will only review it to test the reasoning and evaluate the decision, in order to safeguard trust, transparency and due process.
14. There is, in principle, therefore, no reason to object to the parties agreeing to have their disputes resolved by a competent machine. We know that 60 million such disputes are resolved on eBay every year, apparently without complaint. 90% of those disputes are resolved, according to eBay, in 24 hours without human intervention.
15. This difference between arbitration and litigation is significant. It will take time for courts to develop AI programmes that can be calibrated, tested, evaluated and authorised to assist, let alone replace, judges in their decision making.
16. I visited Shenzhen Intermediate Peoples’ Court in September 2026. There, an AI-powered platform has been developed that allows AI to assist the judge by distilling the important points to be derived from the evidence, listing the issues that the machine thinks need to be decided in order to determine the case, and providing drafts of the reasons and orders that the judge might want to provide to explain and encapsulate their decision. But that last part of the process is based on the judge having decided those issues first. Nonetheless, one judge told us that she could, using the system, do 6 cases in the time she had previously taken to do one.
17. The White Paper that the Shenzhen Intermediate Peoples’ Court issued at the end of September entitled AI in the Judiciary – the Shenzhen Practice” includes the following informative passage:
Judges no longer leaf back and forth between documents and systems to establish the facts of a case. Review becomes a matter of checking, correcting, and supplementing, and work that took a full day now takes tens of minutes.
By taking over formatting, data entry, and proofreading, the System frees judges from mechanical work for the contested heart of the case. Formally, documents are produced strictly to standard, with fewer human errors and defects. Substantively, the reasoning is fuller and answers each contested point. As one lawyer reported: “My client read the reasoning in the judgment and said he felt taken seriously. As the losing party, his sense of grievance was noticeably eased.” The System has now assisted in drafting 678,000 judicial documents.
18. The reality is moving fast in courts, but the consensual nature of arbitration will enable it to move faster. It is that which I shall now explore.
Can we see now how the arbitral process may change?
19. Many of those involved on a daily basis with international arbitrations claim that their cases are so complex and the tensions so great that AI is not likely to undermine the basic process any time soon.
20. Of course, they acknowledge the use of AI programmes such as Harvey and Legora for legal research, and the use of LLMs for the creation of useful summaries and chronologies, but they doubt whether their world will be significantly transformed in the near to medium term.
21. My perspective is different, particularly having seen the Shenzhen model in operation. Complexity does not insulate the arbitral process from change. AI is already being used at every stage of the legal process and that is unlikely to change whatever regulation may be put in place. Clients use LLMs before they come near a lawyer. Lawyers use LLMs in their case preparation, legal research and advice preparation, and arbitrators will use LLMs to process and summarise large volumes of material, if nothing more.
22. The reasons why AI is likely to become integrated into every aspect of arbitral dispute resolution are essentially economic. The ever-increasing complexity of international disputes, which I have already mentioned, has made international arbitration more and more expensive. Parties, even state parties, will not readily accept those costs without question when technology can reduce them. The market will, I believe, become increasingly competitive, with those utilising technology most effectively to cut through complexity eventually winning out.
23. Many lawyers yearn for rules that will prevent, for example, legal submissions or arbitral decisions being created by AI. They suggest that AI should be used to check whether aspects of submissions or an award have been written by a machine. This is, I believe, a hopeless avenue.
24. The one thing that will not change is that, where the parties appoint human arbitrators, those human arbitrators will be responsible for their decisions. It does not matter whether they have been created with the assistance of machines. In my world, I regularly receive material from human judicial assistants. If I utilise that material in my judgments (which I have actually hardly ever done) I am wholly responsible for it. There is no real difference with an LLM. If arbitrators use LLMs, that in no way relieves them of responsibility for every word and figure in their award.
25. The big question is whether the ICDR/AAA machine made decision-making will catch on. That institution has told me that it has started slowly. I expect it will accelerate, when parties realise that it is far more economical than human decision-making and probably no less reliable. There will be studies demonstrating the reliability or otherwise of machine-made decisions and comparing them to the reliability of those made by humans.
26. Once parties have experience of machine-made decision-making for their smaller claims, they will start to ask why it cannot be used for larger claims. Moreover, even if the models used to resolve claims take time to evolve, the parties on both sides of an arbitrated dispute will use those machines to tell them what the outcome is likely to be. As they are both advised of these outcomes, the parties will feel economic pressure to settle at the level predicted by the machine rather than waiting for a decision made – whether or not that decision is guided by the same machines or programmes that the parties are using.
So what about enforcement?
27. For many years, we have hailed the importance of the New York Convention.[3] I am concerned that, when arbitral decisions are made by machines, both consensually and in fact, significant issues may start to arise.
28. Article V(2)(b) of the New York Convention[4] imports a public policy exception to enforcement. It provides as follows:
Recognition and enforcement of an arbitral award may also be refused if the competent authority in the country where recognition and enforcement is sought finds that: … (b) The recognition or enforcement of the award would be contrary to the public policy of that country.
29. Countries around the globe are moving at very different speeds in the adoption of AI. My Shenzhen example is a case in point, but extensive use of AI is being made also in Brazil, Argentina, South Korea and Kazakhstan, to name but a few.
30. It seems likely that parties, against whom enforcement is sought in countries that try to impose harsher regulation on AI usage, could seek to argue that the recognition or enforcement of an award in which AI has played a significant role, would be contrary to that country’s public policy.
31. It would be unfortunate if a hole were ever to be created in the machinery of the New York Convention.
32. The solution, as I see it, is what I have been suggesting for some time. An international debate on good practice in the use of AI in judicial and arbitral decision making is required. SIFoCC, whose steering group I now chair, is attempting to distil good practice in the use of AI in commercial decision-making from across its 71 jurisdictions. It is attempting to identify uncontroversial use cases for AI, and to say where caution is required because of the risk of inadvertently reducing judicial and arbitrator involvement in decision-making. SIFoCC’s project may end up in developing model rules for the use of AI. I think such Model Rules would be valuable to countries at each end of the spectrum I have described.
Conclusions
33. It would not have been appropriate for me to comment on the effect of global conflicts or economic fragmentation on international dispute resolution. But I believe that AI and technological advances beyond it will have an equally fundamental effect on the way disputes arise and are resolved in the modern ever-changing world.
34. It is very important, I think, for international organisations to work together, so that two-speeds do not emerge in international dispute resolution, leading to recognition and enforcement issues.
35. Many organisations, beyond SIFoCC are working on the problem. They include UNESCO, UNCITRAL and the European Law Institute (of which I am first Vice President). I would urge the ICC also to work alongside those I have mentioned and national governments to be part of the solution.
[1] For example, International Chamber of Commerce (ICC), London Court of International Arbitration (LCIA), Permanent Court of Arbitration (PCA), Stockholm Chamber of Commerce Arbitration Institute (SCC), Swiss Chambers’ Arbitration Institution (SCAI), The International Centre for Dispute Resolution (ICDR), in the international branch of the American Arbitration Association (AAA), Singapore International Arbitration Centre (SIAC), Hong Kong International Arbitration Centre (HKIAC), China International Economic and Trade Arbitration Commission (CIETAC), and the Arbitration Center of the World Intellectual Property Organization (WIPO).
[2] https://www.adr.org/ai-arbitrator/
[3] Article V(2)(b) of the Convention on the Recognition and Enforcement of Foreign Arbitral Awards (New York 1958).
[4] Article V(2)(b) of the Convention on the Recognition and Enforcement of Foreign Arbitral Awards (New York 1958).
The post Speech by the Master of the Rolls: The future of cross border dispute resolution appeared first on Courts and Tribunals Judiciary.