Kluwer Arbitration Blog

The short answer to the title of this post is that while a few institutions have already begun to employ artificial intelligence (“AI”), most are in a wait-and-see-and-explore mode as they consider whether and how to make use of it. Most of the current institutional interest in AI is in areas that law and technology guru, Richard Susskind, describes in his book Tomorrow’s Lawyers (at pp. 39-40; see also p. 49) as efficiency initiatives, i.e., making current workflows quicker and lower in cost.

As to using AI in entirely new products or services to resolve disputes, several institutions are giving this thought but there is no clear consensus as to whether these will be introduced in the future, what these new dispute tools may look like, or when they may be deployed.

 

The Survey of Arbitral Institutions for the ICCA Congress Panel on AI

In preparing for the 8 May 2024 session, “International Arbitration: An AI Endeavor” at the ICCA Congress in Hong Kong, panel members sought to capture the state of play of AI’s use in practice, including legal tools enhanced by AI that are currently, or soon will be, on the market. At the same time, the panel queried whether arbitral institutions are already implementing AI. For the latter inquiry, the authors sent a short survey to international and regional arbitral institutions asking them about the current or possible use of AI in four areas:

• Internal processes
• Case management
• Enhancement of current dispute resolution services
• New products and services incorporating AI

A copy of the survey questions can be found here.

Of 18 leading international and regional institutions that received the survey, 11 provided substantive responses; the other 7 either declined to participate or did not respond within the time limit provided. When requesting information from the institutions, we indicated that we would not be identifying respondents by name in our report, only the aggregate results,¹⁾A few of the responding institutions requested that their identity be shared publicly with their data. While we are maintaining our commitment not to publish names, we encourage any of the institutions to share their responses. as shown in the chart below.

 

A Few Are Using AI, While Others Are Assessing Future Deployment

Out of the 11 institutions that responded, 4 indicated they have implemented AI in some form, and all of the responding institutions said they recognize AI’s potential and are considering whether and how to adopt it in the future. Some institutions indicated a preference to proceed with caution via an incremental approach. One institution, for example, is implementing a three-step strategy—to learn “as much as possible” about AI, to determine its best possible use cases, and to identify and mitigate any risks.

 

The Early Adopters

Of the four that have currently adopted AI tools, one employs more than half a dozen engineers and coders to support internal processes, case management, and the institution’s current dispute resolution services. Specifically, AI is used in the preparation of casework reports, drafting of correspondence to the parties and arbitrators, tracking of case progress and compliance with procedural deadlines, assistance with hearing logistics/preparation, and scrutiny of draft awards and procedural orders. Uniquely among the respondents to the survey, this institution also employs AI to create auto-generated portions of arbitral awards, including the case’s procedural history, that are shared with arbitral tribunals and parties.

Another institution has an AI working group to encourage innovation and promote responsible use of AI tools among its employees. The institution reports having identified over 60 potential use cases and says it is working to develop a range of tools to improve all aspects of its dispute resolution services. For case management, this institution has developed an “automated scheduling order tool” for fast-track construction cases. It can prepare a preliminary procedural timetable “in less than a minute” based on a preliminary hearing transcript. The institution says it is currently developing an AI-powered chat bot to assist arbitration filings, and separately, a form of personalized assistance to users in accessing its arbitration rules.

The third institution in this group employs ChatGPT and DeepL Write to enhance its internal processes: to generate ideas for presentations, speeches, marketing materials, memoranda and emails. The institution says it is contemplating the use of AI in other areas.

The fourth institution also uses generative AI tools to optimize its internal processes, principally to brainstorm ideas and produce internal materials more efficiently. That institution has also recently issued a set of rules for AI-related disputes in anticipation of disputes that may arise as AI products become more widespread.

 

Improving Efficiency, Not Revolutionizing Dispute Resolution

Across all institutions that are not presently employing AI, but are considering it, interest is highest in efficiency initiatives for internal processes and case management, i.e., means of improving current workflows by doing them faster and at lower cost. Specific examples include using AI to help track time and costs of proceedings, prepare casework reports, assist with knowledge management and record keeping, and prepare drafts of correspondence to the parties and arbitrators.

The areas mentioned by institutions for the potential use of AI were:

• assistance with arbitrator selection, vetting and conflict checks;
• scrutiny of draft awards and procedural orders; and
• preparation of case summaries and/or identification of key issues for arbitrators.

When it came to inventing new products or dispute resolution services based on AI, there was no consensus on what such products or services might look like. Two institutions indicated that they are not even contemplating this step at the moment. Among those that said they are considering the possibility of new products or services based on AI, there was skepticism and hesitation over whether AI is sufficiently developed to permit this.

 

Conclusion

While a few arbitral institutions have already embraced AI or have developments in progress, it would be inaccurate to infer from the results that those that have not (yet) adopted any AI tools are behind or would find it impossible to catch up. Where AI is being employed by arbitral institutions, in most cases, it is through the use of off-the-shelf tools, not substantial investment into bespoke software.

For example, one institution indicated it is actively considering the adoption of one well-known business software that incorporates generative AI for use in transcription, drafting of notes, creation of action points and summarization functions (among others) to optimize internal processes. Off-the-shelf tools provide a low barrier of entry in the adoption of AI, which may make it easier for other institutions to follow.

There was one point on which all arbitral institutions agreed: AI has the potential to improve their operations and services.

________________________

To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.