SIAC Symposium 2026 Part 2: Artificial Intelligence and the Business of Resolving Disputes

The afternoon of the Singapore International Arbitration Centre (“SIAC”) Symposium 2026 saw delegates split into five concurrent breakout sessions, covering disputes involving China, India and the United States; the energy transition and infrastructure disputes in Southeast Asia; supply chain disruptions and force majeure; Middle East transactions; and legal technology. This post covers the session on legal technology, organised by Young SIAC (“YSIAC”), and the closing plenary with in-house counsel. One question ran through both sessions: although legal work gets better or faster, does the client end up with something it actually values?

Who Moved My Cheese? How Legal Tech is Changing Your Arbitration Practice

Mr Aditya Singh (Member, YSIAC Council; Partner, White & Case) moderated the panel featuring Mr Oliver Armas (Member, SIAC Court of Arbitration; Global Head of International Arbitration, Hogan Lovells Cadwalader), Mr Alexander Fenner (Regional Counsel, Southeast Asia and India, The Boeing Company), Mr Ng Jern-Fei KC (Advocate, Duxton Hill Chambers (Singapore)), and Ms Michele Sonen (Head of APAC & MENA, Jus Mundi). An opening show of hands revealed that much of the room already used generative artificial intelligence (“AI”) in some form, which rather settled the question posed by the session title.

The panel discussed common uses of AI, including legal research across languages, background work on prospective arbitrators and early case assessment. The last is worth a word: a proper first look at the merits was often not worth paying for because the work was out of proportion to what the client needed at that stage, but the tools now bring it within reach. It was also observed that AI could generate gains in document review and practice administration as much as in research. Smaller practices could take on labour-intensive disputes they might once have declined, although this would not necessarily give them a cost advantage over larger firms.

Productivity Gains, and Who Gets Them

It was observed that AI had not made legal teams noticeably smaller, but had changed what a team of the same size could do, including bringing partners into the substance of a case earlier than before. The panel also discussed how firms that had spent money on technology and training would need to talk to clients about how the gains should be shared.

One of the panellists observed that, if offered the same service at a lower price or a better service at the same price, the client would take the former. The discussion also acknowledged the tension in asking for savings while still expecting counsel to check everything the tools produce and to answer for it. It was suggested that this should be addressed at the start of the engagement.

The panel did not entirely resolve the question of how best to capitalise on productivity gains. An hour saved can come off the invoice, go into further analysis or buy senior attention on the file earlier than usual. In the author’s view, counsel should agree with the client at the outset how those gains will be reflected in the service and its price.

Training raised a similar issue. One of the panellists was optimistic: with the mechanical work taken care of, juniors could spend more time reading the record and preparing for cross-examination. A caution was also raised: routine work is where juniors have traditionally learned, and if that training disappears, senior lawyers may have to fill the gap themselves. Using AI tools does not guarantee that the hours freed up will go into better training.

AI in the Hands of the Tribunal

On tribunals' use of AI, the panel was divided, though less than it first appeared. One panellist favoured disclosure, with institutions issuing guidance and the parties settling the detail between themselves, while another panellist saw no good reason why institutions should not simply require disclosure, leaving the parties to decide which uses they would permit. The difference was about who imposes the obligation. Neither doubted that transparency was the right starting point.

Another view was that the line should be drawn at the decision itself: the tribunal may use the tools, but it must not delegate decision-making. It was questioned how clear that line is: a summary of the evidence already involves choices about what to include, leave out and emphasise. The panel discussed whether the answer would differ from case to case and should be discussed properly at the outset. One proposal was to set out in Procedural Order No. 1 (“PO1”) whether, and to what extent, AI would be used, prompting a discussion between the parties and the tribunal about appropriate limits. However, what parties might accept in a dispute over the construction of a clause may not be what they would accept in a case turning on contested allegations of dishonesty. A separate warning was raised about confidentiality and what happens to material once it is fed into an AI system.

It was emphasised that the question was not only whether a human made the decision, but whether a human understood the underlying evidence. The assurance parties would seek depended on the case, including the time and cost of obtaining that assurance from a human. In the author’s view, these observations suggest that simply reserving the final decision to the arbitrators does not adequately address AI’s influence on the decision-making process. An AI-generated summary may shape the tribunal’s understanding of the evidence before it reaches that decision. PO1 should therefore identify the tasks for which AI may be used and specify how the tribunal will review the resulting material, with the extent of that review tailored to the dispute.

Global Disputes, Real Decisions: What In-House Counsel Want from Arbitration Today

Mr Ng Kim Beng (Member, SIAC Court of Arbitration; Managing Partner, Rajah & Tann Singapore LLP) moderated the second and final plenary, with Ms Audra Balasingam (General Counsel, StarHub Group), Mr Damien J. Eastman (retired Deputy General Counsel, Asian Development Bank), Mr Jordan Gimbel (Associate General Counsel (Open Innovation), Microsoft), Mr John Ho (Global Head, Legal, Financial Markets, Standard Chartered Bank), and Ms Alison Lee (Senior Legal Counsel, APAC, FedEx). With telecommunications, development finance, technology, banking and logistics all represented, the discussion showed a diversity of perspectives and that no single feature of arbitration is valued by every user.

What Counts as Success, and Has This Evolved?

The panel observed that the choice of dispute resolution mechanism is made at the negotiating table, where the clause is often the last thing anyone looks at. Confidentiality, neutrality, and enforceability were listed as the reasons a bank chooses arbitration in cross-border transactions. The discussion also noted why arbitration matters even more to multilateral development banks, whose institutional immunities restrict recourse to national courts: a credible dispute resolution process is part of how they stay accountable to their counterparties.

One of the panellists qualified the usual praise of confidentiality. Drawing on past experience, the panellist described cases in which litigating in public against bad actors was the point, because users of the business’s service needed to see the business stand up for them. The forum followed the business objective, not the other way round.

Those objectives can change while a case is on foot. It was observed that rapid changes to an online platform’s products and services could make an ongoing dispute with a developer less important to the business, prompting it to reassess its approach to the proceedings. Another panellist recommended going back regularly to the merits of the case, the management time being consumed and the commercial reason for carrying on with an arbitration. Starting an arbitration should not create a presumption that it has to run to an award. Winning eventually is little consolation if the business has stopped caring about the subject matter.

Procedure in the Service of Settlement

The panellists had concrete reasons for using the flexibility provided for in the SIAC Rules 2025. Preliminary determination was welcomed, with a dispute over a contractual liability cap being given as an example. Deciding early whether the cap applies can close much of the gap between the parties' views on quantum, and once the gap is smaller, a deal becomes possible. An early ruling on one issue can therefore move the parties towards settlement even if it does not dispose of the case.

The panel observed that a logistics business cannot simply replace its delivery partners, ground handlers or airport operators, so preserving the relationship matters in any dispute with them. The SIAC–SIMC Arb-Med-Arb Protocol gives the parties a structured pause in which to look for a commercial solution, with any settlement recorded as a consent award. A similar point arose from the banking side: a recoverable settlement that keeps the relationship intact is worth more than a larger award that cannot be enforced.

One panellist questioned claims about arbitration's cost-effectiveness. The panellist recalled a modest dispute run under procedures wholly out of proportion to its value and expected external counsel to advise how revised institutional rules could improve arrangements in existing contracts. Another panellist observed that costs were also borne in operational terms: a wide document request, for example, could pull staff away from their daily work.

Efficiency also counts for little if it stops at the award. The panel cautioned that a faster procedure does not rule out the risk of delays in local courts at the enforcement stage, and that speed cannot come at the expense of a proper opportunity to be heard. The procedural choices made at the outset have to fit both the nature of the dispute and the place where the money will eventually be collected.

If there is a lesson for counsel in all this, it is a simple one: before proposing the next procedural step, ask again what the client is trying to achieve. An early ruling may be worth having because it makes settlement possible, and a stay in favour of mediation may be worth taking because the parties still have to trade with one another afterwards. The flexibility in the arbitral rules achieves little if the case is simply run to the standard timetable without anyone asking those questions.

Closing Reflections

Mr Vivekananda Neelakantan (Registrar, SIAC) closed the day by marking the institution's 35th anniversary and thanking the community that has grown up around it.

The afternoon's discussions set a practical test for what comes next. Clients want to see how savings from technology show up in their own case, and be able to trust the judgment exercised over it. They also want counsel prepared to ask whether the fight is still worth having. The technological and procedural improvements discussed at the SIAC Symposium 2026 will matter most when they change day-to-day business decisions.