The Chartered Institute of Arbitrators’ new Guideline on the Use of AI in Arbitration

Global efforts to regulate the use of AI in legal proceedings

There has been some effort to regulate the use of AI in legal proceedings. The European Union recently enacted the Artificial Intelligence Act (Regulation (EU) 2024/1689), being the first comprehensive horizontal legal framework for the regulation of AI systems across the EU. The Act classifies AI according to its risk level, places obligations on providers and users of high-risk AI systems and specifies requirements with which General Purpose AI model providers must comply. AI systems that are intended to be used by a judicial authority, or to assist judicial authorities, are considered high-risk AI systems and therefore, providers and users of these systems must comply with the Act. Various judiciaries including the UK Courts and Tribunals Judiciary, Colombia (which adopted UNESCO’s Guidelines for AI Use in Judicial Systems) and the Californian courts have developed guidance for practitioners on the use of AI systems in courts and tribunals. Professional organizations including the Silicon Valley Arbitration & Mediation Center, the Law Society of New South Wales and the American Bar Association have also published guidelines for practitioners on the responsible use of AI in dispute resolution proceedings.

Alongside the above, CIArb’s practical guidelines on the use of AI in arbitral proceedings seek to “give guidance on the use of AI in a manner that allows dispute resolvers, parties, their representatives, and other participants to take advantage of the benefits of AI, while supporting practical efforts to mitigate some of the risk to the integrity of the process, any party’s procedural rights, and the enforceability of any ensuing award or settlement agreement.” The Guidelines are an important resource for practitioners and arbitrators despite their non-binding nature and parties may choose to incorporate these Guidelines or the template agreements provided into the arbitration by agreement.

The Guidelines on the use of AI in arbitration

The key principles can be summarized as follows:

• Arbitrators’ powers to give directions and make procedural rulings on the use of AI: The use of AI falls within the general power of arbitrators to conduct proceedings, including giving directions and making procedural rulings. Tribunals are encouraged to record decisions around AI and address any contentious use of AI in their awards.
• Party autonomy: Where the Parties (1) have not raised the issue in their initial communications with the institution or the arbitrators or (2) the arbitration agreement is silent or unclear on the use of AI, arbitrators are encouraged to invite the Parties to express their views on the use of AI in the proceedings at the appropriate time.
• Ruling on use of AI and admissibility of AI-generated material in the arbitration record: Arbitrators may make a ruling on the use of AI in arbitral proceedings or admissibility of evidence where there is disagreement between the Parties or where the use (or non-use) of AI by one or more party jeopardizes the integrity of the arbitral proceedings. In making a ruling on the use of AI, arbitrators are encouraged to take into consideration any benefits, and consider the nature and specific features of the AI tool.
• Disclosure: Disclosure may be required where use of an AI tool has an impact on (1) the evidence, (2) the outcome of the arbitration or (3) otherwise involves a delegation of an express duty toward the arbitrators or any other party.
• Discretion over use of AI by arbitrators: Arbitrators may consider using AI tools in the context of their mandate to enhance the arbitral process including its efficiency and the quality of decision-making.
• Transparency over use of AI by arbitrators: Arbitrators are encouraged to consult with the parties on the use of any AI tool used by the arbitrators.
• Template AI Agreement: Appendix A is a template Agreement to regulate the use of AI to arbitration proceedings. In particular, this extends the scope of the agreement to the parties, their representatives and their experts.
• Template Procedural Order: Appendix B is a template Procedural Order on the use of AI in proceedings. This places additional duties on the Tribunal to ensure that the integrity of the proceedings and enforceability of the award are preserved against cost concerns and allows specific guidelines to be put into place for ‘High Risk AI Tools.’

Practical perspective: The application of AI tools in arbitration

The Guidelines consider the benefits and risks of using AI tools for improving legal research, expediting data analysis, streamlining the collection of evidence and generating hearing transcripts against the risks of the use of AI tools in arbitral proceedings.

Some of the potential risks of the use of AI in arbitral proceedings include loss of confidentiality, data integrity, cybersecurity, algorithmic bias and broadening the grounds to challenge an award, including failure to follow due process (for example, by using AI in a matter outside the scope of the procedural order) which could potentially increase the frequency of challenges to award.

The Guidelines also have the potential to cause the parties to incur additional costs by introducing an additional arena for the parties to do battle in, as parties explore the extent to which AI can be used in arbitral proceedings. While these Guidelines may result in more careful and considered use of AI tools, as well as potential cost savings using AI itself, the Guidelines may to lead to additional costs should differences arise about how AI tools should be used.

Experience of AI tools in dispute proceedings

The Guidelines highlight the need to rely on legal providers who are experienced in extracting value from AI tools, while at the same time maintaining appropriate safeguards in the conduct of arbitral proceedings. This firm has successfully used AI in appropriate circumstances to enhance the delivery of legal services for clients in a variety of scenarios:

1. Using business intelligence tools and machine learning in a dispute over uncollected VAT.
2. Using a combination of defined terms, machine learning and predictive coding through a dedicated document review platform in a high-value and complex dispute where our client had to consider disclosure in respect of over 400,000 documents.

The firm has developed guidelines on the responsible use of AI which acknowledge the ethical, legal and social challenges but embrace the opportunities presented by AI to deliver high-quality, innovative and client-focused legal services.