Justice Philip Jeyaretnam: Keynote address at the 11th International Chamber of Commerce (ICC) Asia Pacific Conference on International Arbitration

Keynote Address

The Relationship Between Specialised Courts and Arbitration in Resolving Industry-Specific Disputes

Friday, 26 June 2026

The Honourable Justice Philip Jeyaretnam

Judge of the High Court

Supreme Court of Singapore

President, Singapore International Commercial Court

I. Introduction

1. It is a pleasure to be here to address such a distinguished gathering of international arbitration practitioners. I congratulate ICC for convening this conference, now in its eleventh year, and for assembling a programme that addresses many of the pressing questions facing international arbitration today – from the debate on dissenting opinions in arbitration, the proper principles for cost awards, to discussions on the revised 2026 ICC Arbitration Rules. That last topic could itself occupy a day, given the number of changes made. Since the inception of this conference series, the volume and value of international arbitration has grown immensely in the Asia Pacific region and in Singapore in particular. This has been accompanied by a massive increase in Asia Pacific – based arbitration counsel and arbitrators. The infusion of new practitioners in our region can seed and propagate new ideas and better practices.

2. My address today concerns the relationship between specialised courts and arbitration in resolving industry-specific disputes. I will make and elaborate on two points. The first is that arbitration’s strengths of finality, confidentiality, and the freedom to appoint arbitrators, come with a limitation. Arbitral awards are confidential and non-precedential; the law cannot develop through arbitration alone.

The second is that this is where courts adapt or are important, especially in the commercial world of business, where there is constant technological and organisational innovation. Courts develop the law through published, reasoned judgments that bind and guide future disputes. I will close by floating a proposition that ties these two points together: whether a right of appeal from arbitral awards on questions of law could improve the system – expert tribunals resolve complex industry-specific disputes, with the courts available to give authoritative and publicly reasoned answers on the novel legal questions such disputes raise.

3. Before I proceed further may I also commend ICC’s approach in not by its rules excluding rights of appeal at the seat. Leaving such options up to parties properly reflects the principle of party autonomy.

II. Courts as Superintendents of Arbitration

4. I begin with the courts’ supervisory function. Courts do not sit in substantive review of arbitral awards. Once an award is made, the only avenues for curial intervention are an application to set it aside or a challenge to its enforcement. Under the UNCITRAL Model Law, which forms the backbone of Singapore’s International Arbitration Act and the arbitration legislation of many jurisdictions, the grounds for setting aside an award are deliberately narrow. They concern jurisdiction, procedural fairness, arbitrability, and public policy.¹ Singapore’s International Arbitration Act includes two further grounds: fraud or corruption in the making of the award, and breach of the rules of natural justice causing prejudice to a party’s rights.²

5. These grounds are not easily made out. For instance, a breach of natural justice requires the challenging party to satisfy four cumulative requirements established in Soh Beng Tee & Co Pte Ltd v Fairmount Development Pte Ltd [2007] SGCA 28 (“Soh Beng Tee”): which rule of natural justice was breached, how it was breached, the connection to the making of the award, and resulting prejudice.³ All four must be established before the court will intervene. The Court of Appeal affirmed this in its recent decision of ONI Global Pte Ltd v GNC Holdings LLC [2026] SGCA(I) 3. The case arose from a franchise dispute between an American health supplement company and its Singapore franchisee, which had terminated the franchise agreements and rebranded 54 stores overnight. The tribunal found in the franchisor’s favour and ordered, among other things, that the store leases be assigned back. In crafting that specific performance order, the tribunal imposed detailed terms on matters such as landlord consent, employee protection, and the timeline for assignment without first inviting the parties’ submissions on those precise terms.

6. The franchisee resisted enforcement of the award on grounds pertaining to public policy, natural justice, and scope of submission to arbitration. The SICC rejected most of these grounds but found that three orders in the award were made in breach of natural justice. The Court of Appeal overturned that. It held that the four requirements in Soh Beng Tee are cumulative, and the SICC had erred by proceeding directly to prejudice without first establishing that any breach had occurred at all.⁴ On the facts, the detailed terms of the order were a reasonably foreseeable outcome of the way the parties had argued their case. The franchisee had itself raised concerns about third-party hardship and landlord consent throughout the arbitration; the tribunal was doing no more than balancing those very equities in its order. Where an outcome flows reasonably from the arguments the parties actually advanced, there is no breach of natural justice.⁵

7. Courts are strict on the limits of curial intervention. The setting aside grounds are not a back door through which a disappointed party may re-litigate a lost award. A party cannot dress up a substantive appeal as a procedural complaint and expect the court to look behind the award. Because there is no appeal on the merits, errors of fact or law in an arbitral award are final and binding on the parties.⁶ There is no avenue to correct such mistakes. Parties who choose arbitration accept, in the language of the Singapore Court of Appeal in Soh Beng Tee, “the attendant risks of having only a very limited right of recourse to the courts”.⁷ This is by design, because arbitration’s value lies in finality, confidentiality, and the freedom of parties to choose their decision-makers.

III. The Advantages of Arbitration

8. That freedom to appoint arbitrators is especially important for disputes in specialised fields. In recent years, construction, engineering and energy disputes account for more than 40% of the cases registered with the ICC.⁸ These are not disputes that yield to legal analysis alone. A complex offshore energy dispute may turn on whether a particular drilling methodology met the applicable technical standard, whether a commissioning delay was caused by a design defect or a failure of project management, and how consequential losses should be allocated across a multi-party contractual chain. A tribunal can be appointed that will approach such a dispute with an understanding of how the industry actually operates. Such industry-specific knowledge cannot be easily replicated by a generalist decision-maker. The same is true across a range of technically demanding fields that generate the bulk of complex international arbitration, such as construction, shipping, energy and pharmaceuticals. Choosing arbitrators with domain knowledge can positively contribute to the quality of the outcome.

9. This is particularly significant in emerging fields. Disputes involving specialised technologies such as digital assets and AI account for around 6 to 8% of arbitrations at institutions like the ICC.⁹ This percentage will only grow in the coming years. In such disputes, the deepest technical understanding is often possessed by younger members of the profession. Arbitration’s freedom of appointment allows parties to go where the expertise is, to select decision-makers whose knowledge matches the dispute. Courts, by contrast, cannot be bespoke in this way.

10. However, it would be wrong to discount the value of courts, which are repositories of knowledge and experience. In the commercial field, this is evident in the commercial lists of the Singapore High Court and in the SICC, which uniquely combines Singapore commercial judges with international commercial judges from both common law and civil law traditions. With experience comes wisdom. Judges also benefit from sustained and systematic judicial education, such as conducted by the Singapore Judicial College. This includes education in new technologies such as digital assets and AI.

11. Moreover, there is a disadvantage to the arbitration model. Arbitral awards are, by default, confidential. Publication requires the consent of the parties, and most parties do not give it. Even where institutions publish awards, they do so in redacted or anonymised form, often years after the event. And whether published or not, awards carry no precedential weight. They do not bind subsequent tribunals. They do not form part of a body of law.

12. The consequence is that arbitration does not develop the law. Disputes in technically complex and commercially novel fields are resolved as though each were the first of its kind. Tribunals in parallel arbitrations involving the same industry and the same legal questions may arrive at entirely inconsistent outcomes, with neither aware of what the other decided, and no mechanism for correction.

13. Contrast this with what courts do. Judgments are published in full, subjected to academic commentary and scrutiny, cited and distinguished in subsequent cases, and refined over time into a coherent body of doctrine. The law develops in public, and it develops for everyone’s benefit. Singapore’s courts have demonstrated this in novel fields. In Quoine Pte Ltd v B2C2 Ltd [2020] SGCA(I) 2, the Court of Appeal grappled for the first time with the legal consequences of algorithmic trading on a cryptocurrency exchange, addressing whether trades executed autonomously by software could be reversed on grounds of unilateral mistake, and what knowledge could be attributed to a computer programme. That decision, and the SICC judgment that preceded it, will continue to shape how exchanges and market participants structure their platforms and contracts. Digital and crypto assets have raised novel legal questions. In ByBit Fintech Ltd v Ho Kai Xin [2023] SGHC 199 (“ByBit”), I had occasion to consider whether cryptocurrency is property capable of being held on trust. I held that it is. Similar decisions have been reached in Australia, where the decision in ByBit has been discussed in, for example, an excellent extra-judicial article by Australian Federal Court Justice Ian Jackman in his paper “Is Cryptocurrency Property?”. Such decisions have practical consequences across a broad range of digital asset issues, from insolvency and custody to the assertion of proprietary rights. They happen when disputes come before courts, and their publicly available reasoning supports the step-by-step development of the common law.

14. When courts resolve questions of law in novel fields, their answers enter the public domain and become available for incremental refinement. Practitioners can advise clients with greater confidence. Parties can structure their transactions against a known legal background. Regulators and legislators can respond to a developed body of doctrine rather than operating in a vacuum. None of this is possible when the same questions are solved, however expertly, behind the closed doors of a confidential arbitration.

IV. Appeals from Arbitral Awards on Questions of Law

15. This brings me to the proposition I wish to float for your consideration today: the utility of providing a right of appeal from arbitral awards on questions of law. I do not suggest a broad merits review that would effectively reduce arbitration to a first instance court proceeding. What I suggest is a mechanism by which parties may, with the leave of court, refer a contested question of law for judicial determination. The concept is not without precedent. Under the Singapore Arbitration Act 2001, which governs domestic arbitration, a party may appeal to the High Court on a question of law arising out of an award, with leave, where the determination of that question will substantially affect the rights of one or more of the parties.¹⁰ The English Arbitration Act 1996 similarly permits such an appeal on a question of law arising from an arbitral award.¹¹

16. Adopting such a mechanism for international arbitration in Singapore would serve two purposes. The first is error correction. Sometimes, the tribunal simply gets the law wrong. Rather than the losing party trying to shoehorn its dissatisfaction into a challenge on natural justice grounds, the error could be addressed directly in an appeal. The second purpose is where novel questions of law arise in arbitration, questions on which reasonable minds may differ and on which there is no settled authority. Parties would have a route to seek a definitive determination by the courts. Published appellate decisions on questions of law arising out of arbitration would generate precedent that benefits not just the immediate parties, but every market participant grappling with the same unsettled issue. There is an additional positive consequence. The availability of an appellate avenue on law may itself improve the quality of arbitral reasoning. If arbitrators know that their legal analysis may be subject to judicial scrutiny and through that to public attention, they will be encouraged to engage carefully with the applicable principles.

17. There are countervailing considerations. Parties choose arbitration for finality and confidentiality, and an appeal mechanism creates the risk of additional time and cost. There is also the fair question of what a court, which may lack the technical expertise of the chosen tribunal, can meaningfully contribute on a specialist point of law.

18. But those objections are answerable. The mechanism I propose is not a general right of appeal but a limited one, available only with leave of court and only where the question of law will substantially affect the rights of the parties. Fact-finding remains entirely with the arbitral tribunal; whether the facts were correctly found or the evidence properly weighed is not open to challenge. An appeal on a question of law does no more than subject the tribunal’s legal reasoning to scrutiny, produce a published judicial decision, and allow the law to develop. Finality over the facts is still preserved.

19. Further, in the Singapore context, court proceedings are fast. For example, challenges under the Model Law are typically disposed of within 6 months at first instance, both in the General Division of the High Court and in the SICC. In Singapore we are mindful of the overall life cycle of a dispute from commencement to enforcement.

20. Turning to confidentiality, courts have power to anonymise parties and redact commercially sensitive material while still publishing the legal reasoning that constitutes the precedent. There need not always be full disclosure of the dispute. As for expertise, the question referred would be one of law, not of technical fact. Courts, particularly specialised courts, are institutionally equipped to resolve novel questions of commercial law. The SICC, with its mixed bench of Singapore commercial judges and international judges drawn from common and civil law traditions, brings precisely the comparative breadth that emerging fields of 8 commerce demand. Where industry-specific context is needed, it can be supplied through expert evidence and submissions from relevant industry bodies.

V. Conclusion

21. I conclude by observing that the disputes that will define the next decade of international arbitration, in digital assets, artificial intelligence and other technologies, are already arriving. Some of these will be resolved in confidential arbitration by expert tribunals. But the legal questions those disputes raise will not go away once the award is made. They will arise again, in the next dispute, and the one after that, resolved each time from scratch, without the benefit of what came before.

22. A limited right of appeal on questions of law arising from arbitration is one way to ensure that the resolution of arbitral disputes helps build a legal foundation for novel areas of commerce. Singapore’s own domestic arbitration legislation already provides a model. We ought to consider whether we are willing to extend such an approach to the transnational sphere.