From Algorithms to Awards
Ciarb Guideline on the Use of AI in Arbitration (2025)
A word from Claire de Morel Westgaver
I became the new Chair of Ciarb’s Technology Thought Leadership Group shortly after the launch of the very first version of Chat GPT in late 2022, which transpired to be a watershed moment for Artificial Intelligence (AI) in law. As the world leading organisation for arbitrator training and certification, Ciarb is particularly well positioned to provide guidance on the use of AI in arbitration and I knew from my discussions with fellow arbitrators that guidelines on the use of AI in arbitration would be welcome by the international arbitration community. I therefore suggested that this be the next project for the Technology Thought Leadership group.
"As the world leading organisation for arbitrator training and certification, Ciarb is particularly well positioned to provide guidance on the use of AI in arbitration"
After a few brainstorming sessions on use cases and risks, the Ciarb AI Guideline Drafting Committee started developing an outline which revolved around arbitrators, in particular arbitrators’ powers to make rulings on the use of AI by parties and the use of AI by arbitrators. After defining the approach we wanted to adopt for each of the sections, different sub-groups started drafting different parts of the guideline. We continued to have working sessions throughout the process to discuss progress and reach consensus on certain issues such as disclosure of the use of AI by parties and by arbitrators.
Due to the speed at which AI technology is developing and the evolving regulatory framework, a guide based on AI tools and use cases did not seem appropriate to the Drafting Committee. There are so many ways in which AI can be used in arbitration and some of these ways are yet to be invented! We felt that we needed to give practitioners the right tools to assess risks and adopt procedures tailored to the state of the AI development and the dispute at hand. Our aim was to provide assistance with navigating the complexities of integrating AI into the arbitral process through guidance that would apply to any AI tool and any context in which such AI tool might be used. This seemed to be the best way to ensure that the guideline would remain relevant for at least a few years after its launch, although members of the Drafting Committee appreciate that it may need to be revisited earlier.
One of the purposes of the AI Guideline is to educate practitioners about potential risks and considerations that may be relevant when adopting AI-related measures for specific cases. Another one is to provide practical tools to be used when implementing any such measures.
Due to the confidential nature of arbitration and the absence of any mandatory disclosure obligation, it is impossible to measure the use of AI in arbitration. However, some surveys have been conducted in the past and these show that AI is already being used on a regular basis and for a variety of tasks. These include document review and document production, legal research, translation, interpretation and transcription services as well as text generation for orders and arbitral awards.
Claire Morel de Westgaver, Chair of the Technology Thought Leadership Group
Claire Morel de Westgaver, Chair of the Technology Thought Leadership Group
The expectation is that parties will increasingly use AI in the coming years and that LegalTech providers will continue to develop and launch new tools. I also expect in-house teams to use AI themselves. It will be interesting to see how this will affect legal training and also ESG-related reporting obligations due to AI’s impact on the environment. Another ramification of the development of AI on arbitration, on which I am advising clients, is litigation risks associated with the deployment of AI tools in their business operations.
Technology permeates all aspects of the economy and transactions are increasingly cross-border, which means that more and more technology disputes are now resolved by arbitration. In my experience, parties and counsel look for arbitrators who can manage cases and understand technology as well. I see a broad range of technology disputes crossing my desk. Breach of contract claims relating to copyright or patent licenses and service agreements are quite common. Technology disputes can also be project-based and involve scope of work and delay issues. They can also arise in the context of an M&A deal where the target is a technology company, for example. I have also seen data breaches and trade secret cases going to arbitration.
"We felt that we needed to give practitioners the right tools to assess risks and adopt procedures tailored to the state of the AI development and the dispute at hand"
What I find particularly fascinating with this area is the combination of AI’s disruptive effect and the governance aspects which involve many components including the technology, AI-related regulations and implications for the administration of justice and the rule of law. Ironically, my work on the use of AI brought me to “re-assess” the nature and purpose of arbitration and reminded me how much international arbitration is a human endeavour.
I am very proud of the end product and the collaborative process that led to it. The Drafting Committee consists of a group of experts who are truly experienced with AI-related issues and with very different backgrounds. Everyone had something different and interesting to contribute to the project which made the discussions extremely rich.
The Guideline was put together by:
Policy and Professional Practice Team
Cristen Bauer, Ciarb Head of Policy
Technology Thought Leadership Group AI Guideline Drafting Committee
Claire Morel de Westgaver, Chair
Amy Endicott
Annabelle Onyefulu ACIArb
Carlos Carvalho MCIArb
Fabio Solimene
Harry Borovick
Karolina Jackowicz FCIArb
Kateryna Honcharenko MCIArb
Matthew Lavy KC
Maud Piers
Meriam Al Rashid
Peter Neumann FCIArb
The AI Guideline outlines the benefits and risks of the use of AI in arbitration, sets out general recommendations on the use of AI in an arbitration, addresses arbitrators’ powers to give directions and make rulings on the use of AI by parties in arbitration, and finally, addresses the use of AI in arbitration by arbitrators.
It also comprises two appendices. Appendix A is a template Agreement on the Use of AI In Arbitration. Appendix B is a template Procedural Order on the Use of AI in Arbitration.
Ciarb's forward-thinking approach
"Practitioners across legal fields are crying out for guidelines which respect the potential advantages of AI, but give them clear frameworks for best practices. The AI Guideline demonstrates Ciarb’s forward-thinking approach – something more regulatory bodies should urgently replicate," explains Harry Borovik, who was part of the drafting committee.
"The AI Guideline should help the ADR industry rapidly adopt AI tools with a clear understanding of appropriate guardrails. A the very least, it will help anyone in ADR understand the appropriate questions to ask when adopting, using or selling AI tools."
"I use AI as a summarisation tool, to refine drafting, and occasionally to help structure unstructured data - though I always review everything carefully to ensure accuracy."
"I expect the changes AI makes to ADR to be gradual — so much so that they may hardly be noticed. Most AI adoption will take place behind the scenes, within law firms and at an arbitrator’s desk. This is precisely why it is so important to establish best practices early on and develop a shared understanding of what is acceptable and what is not," says Carlos Carvalho MCIArb.
"I use AI as a summarisation tool, to refine drafting, and occasionally to help structure unstructured data - though I always review everything carefully to ensure accuracy. However, I still rely heavily on my own research and do not use AI for substantive legal analysis. I have also seen AI used for transcription, but this still requires human verification and editing to ensure reliability."
"Guidelines like this one create space for greater efficiency while upholding clear standards for disclosure, transparency, and human oversight"
"The Guideline is broad and accessible, providing practical guidance and raising awareness among all key actors in arbitration. I view it as both a reference point and a conversation starter, outlining principles that inform decisions and policies while serving as a foundation for developing context-specific rules, all while preserving room for professional judgement," says Prof. Dr Maud Piers.
"Guidelines like this one create space for greater efficiency while upholding clear standards for disclosure, transparency, and human oversight. The framework they provide is grounded in legal ethics and arbitral principles, enabling progress while ensuring that professional judgement remains central. Crucially, they do not seek to replace human judgement, but rather to support and structure it," she continues.
The ADR ecosystem
by Ciarb President Prof. Dr Mohamed S. Abdel Wahab C.Arb FCIArb
The international arbitration ecosystem is experiencing colossal challenges and is undergoing significant changes resulting from the political, economic, social, environmental and technological trends prevailing globally. However, challenges and changes bring about opportunities and prospects, and AI is no exception.
Given Ciarb's vision to pioneer the future of dispute resolution through innovative, efficient and fair practices worldwide, and in fulfilment of its role as a global thought leader and standard setter in ADR, Ciarb's Guideline on the Use of AI in Arbitration comes as a welcomed addition to the impressive assemblage of guidelines that the Ciarb has gifted to the global ADR community."
"The paradigm shift caused by the disruptive effect of AI and the integration thereof in arbitral proceedings is paving the way for a more dynamic and tech-driven future, and the Guideline is both timely and momentous.
In four parts and two appendices, the Guideline offers innovative standards, recommendations and practical guidance to all stakeholders and users of international arbitration on topical issues relating to the benefits and risks of the use of AI in arbitration, the parameters for disclosure of use of AI tools, and the arbitrators’ powers to give directions and make rulings on the use of AI in arbitration to safeguard the integrity of the process, observe due process considerations and ensure the enforceability of arbitral awards and/or settlement agreements.
The template agreement and template procedural order on the use of AI in arbitration as appended to the Guideline will no doubt prove valuable to all parties, arbitrators and counsel involved in arbitral proceedings across the globe.
Many thanks to the Drafting Committee for their work and time and to the global ADR community for their confidence and use of Ciarb's suite of Guidelines.
The future of AI
by Matthew Lavy KC
As an ADR practitioner, the thing I find most exciting about being in the tech space is the technology. I love the way technology has the power to transform lives and the way the world works.
Over the next couple of years we are going to get an increase in adoption of LegalTech. We're no longer going to be looking at LegalTech in labs: it's going to be part of everyone's practice. It will dominate for a while. Inevitably, when you have a new piece of technology (particularly one like AI) people need to be trained to use it and to incorporate it into their workflows. Therefore, it's going to dominate practice.
We're going to find ourselves going through a period where a lot of brainpower will be spent on thinking, “OK, how do I use AI for this?” It won’t last long because once there's adoption, then AI will just be part of our practice. It won't dominate any more than, say, a word processor does today.
AI will be part of the fabric of our professional lives. I think we'll see the same with LegalTech generally and AI in particular. It won't be at the forefront of people's minds. It will become totally unremarkable, which is how it should be. It's a tool at the end of the day, like Google.
Built with Shorthand