International Arbitration Trends in 2026
The integration of AI in international arbitration has moved from novelty to necessity. While its promise of increased efficiency and cost-effectiveness is clear, the expanded use of AI tools for document analysis, drafting, research and more is creating a new species of procedural challenges.
Safeguarding the integrity of the arbitral process now requires further attention to transparency, data security and the “human in the loop” requirement for decision-making.
With no unified regulatory framework yet in place, arbitrators, counsel and parties are left to navigate a growing patchwork of institutional guidelines.
In detail
The dicourse dilemma
Disclosure obligations generally ensure the integrity of the arbitral process. While there is a trend towards disclosure of AI use, the scope of disclosure remains a point of debate. For example, institutions differ in their approach:
- The broad view (2025 AAA-ICDR Guidance on Arbitrators’ Use of AI Tools): The AAA Guidance encourages disclosure where AI tools materially impact the arbitration process or the arbitrators’ reasoning. However, the guidance leaves open the precise content of the disclosure obligation.
- The targeted view (2024 Silicon Valley Arbitration & Mediation Center Guidelines on the Use of Artificial Intelligence in Arbitration): The SVAMC guidelines suggest that general AI use does not require disclosure. Where appropriate, however, they suggest that the following details may help reproduce or evaluate the output of an AI tool: (1) the name, version and relevant settings of the tool used; (2) a short description of how the tool was used; and (3) the complete prompt and associated output.
In practice, parties and counsel will likely resist disclosing the “internal” use of AI (e.g., specialized legal AI), arguing it is no different from using a search engine, provided that the tool operates within a secure environment and does not itself create a risk of confidentiality or data breaches. The challenge for 2026 is standardizing the “materiality threshold”, i.e., the point at which AI’s role moves from efficiency aid to substantive contributor.
The central challenge of integrating AI in arbitration is balancing its promise of efficiency against the need to safeguard procedural fairness, transparency and the legitimacy of the arbitral process.
Rohit Bhat
Partner
A critical procedural risk is the submission of AI-generated “hallucinations” — fictional case law or fabricated evidence. To some extent, the occurrence of hallucinations can be minimized through more precise prompting by users as they become more adept at using AI tools. At the same time, providers of AI tools are taking steps to further mitigate the risk of hallucinations. Even so, with an increasing number of documented cases globally of AI-generated fake citations, tribunals may consider imposing a duty of human verification, i.e., requiring parties to certify that a human has reviewed all submissions for factual and legal accuracy. Such measures would align AI‑assisted drafting with existing duties of candor and could be reflected in early procedural orders or soft law instruments.
Practical takeaways
AI is already reshaping the conduct of international arbitration, and its procedural implications cannot be ignored. To stay ahead, tribunals, parties and counsel should consider these practical steps:
- Address AI upfront: Proactively address disclosure and verification early in the proceedings, ideally from Procedural Order No. 1 onwards.
- Certify accuracy: Ensure any AI-generated materials are reviewed and certified by a qualified human, minimizing the risk of errors or hallucinations.
- Monitor evolving guidance and best practice: Emerging institutional guidance offers useful building blocks, but practice will develop case by case.
- Educate your team: Ensure your arbitration and legal teams are trained on both the capabilities and risks of AI in proceedings.
The challenge for 2026 and beyond is to ensure that AI enhances efficiency without compromising fairness, transparency or the legitimacy of the arbitral process. These principles are embedded in the way we work. Please contact us to discuss how we can support you.
Authors
