The adaptation and incorporation of AI tools in various professional and non-professional spheres is a common phenomenon now. The user friendly and readily available interface is giving a significant push to their demands in the marketplace. Acknowledging this fact, it can safely be said that the legal fraternity is also not untouched from this technological boom.
This article aims to analyse the amalgamation of arbitration industry with AI in the wake of the ICCA Congress at Hongkong’24 themed: “International Arbitration: A Human Endeavor”.The conference was held from 5th May 2024 to 8th May 2024 highlighting and inducing discussions about several evolving aspects of International Arbitration. On 8th May 2024, it hosted the panel discussion titled “International Arbitration: An AI Endeavour”. It aimed at addressing the interplay of AI in legal practice, including the legal research tools which are currently or soon to be available in the markets.
The usages of AI in legal practices are not unheard of, however its “unregulated” assistance in official documentation and processes, subsequently resulting in bogus authorities, breach of party information etc. has become a thing of concern now. Some attempts however, at regulating the same will also find mention in this article proceeding forward.
Arbitration with AI
Generative AI tools are trained in large language models which allows them to swiftly search across various databases and quickly extract the insights after drawing a comparison and summarizing the same. It has significantly reduced the input requirement of arduous manual labor in document analysis; like searching and indexing for facts in a testimony, or identifying discrepancies in transcripts or creating annexures and timelines after summarizing the bulky mass of documents.
The aid in text drafting with the use of tools like ChatGPT or Claude or LinerAI enabling a rapid crafting of professionally customized drafts, is also significant. Eliminating the repetitive aspects of the arguments, or creating customized drafts for case management according to the specifics of a case at hand, or possibly drafting initial arbitral awards based on the merits of the case, etc are some of the ways in which assistance is available at disposal.
Some of the AI tools which are being used in the legal industry are Casetext for legal research, ChatGPT for drafting and customization, Claude for practical applications, Everlaw, Lex Machine, LawGeex, Due Diligence and many more.
With its power to examine large data sets, AI is said to be more accurate and less biased than human decision-making. However, AI is devoid of emotional intelligence and may miss the fine details. It is assumed here that this may affect its ability to yield “fair” or “merit-based” outcomes. The likelihood is a strict interpretation of “correct” legal outcomes, not necessarily what is best for the parties involved, meaning that this topic is arguable. One such case is how much arbitrators should rely on a strict interpretation of the law to resolve legal matters.
HOWEVER, with the unceasing evolutions in technology, it seems just to be matter of time before this objection is countered by the AI General Intelligence.
ICCA Hongkong’24 survey results
In the survey, it was sought to capture whether the arbitral institutions were taking AI assistance already or not. The questions for the survey were sent out to the regional and international institutions dealing in arbitration, out of which 11 responded. The questionnaire probed into the implementation of AI in (accessible here ):
- Internal Processes
- Case Management
- Enhancement of current dispute resolution service
- New Products and services incorporating AI
4 out of the 11 institutions responded in an agreement to the use of AI in one or the other form. One of them has employed a team of coders and engineers to assist in the case management, dispute resolution services and internal processes. AI assistance is being used for drafting arbitral awards, documents for both parties and arbitrators, case briefings with the use of logistics and many more.
The second institution is promoting informed application of AI tools and has formulated an AI specific working group. Over sixty possible areas of its use have been pinned. An “automated scheduling order tool” has been developed for creating quick preliminary timetables of procedures involved in fast-track cases based on the initial hearing transcripts.
The third institution has been engaging in the use of ChatGPT and DeepL for refined transcripts with regard to the internal processes for emails, speeches, presentations etc. The implementation of AI in other fields is under close scrutiny.
Last but not the least, the fourth institution has also been using AI for internal processes to produce materials and also has remarkably, issued a set of guidelines for AI related disputes.
However, all 11 of them acknowledge the potential of AI and are considering its integration with the professional work in one or the other ways. They collectively agreed to its flair for improving the work efficiency. The rest 7 who haven’t yet employed AI, are interested in deploying the same for case managements and internal processes.
Regulation Attempts
Some attempts have been made for regulating the unbridled use of AI in the official legal work and prescribing a set of guidelines for the same, to make it a bit streamlined and systematized. This endeavor has been necessitated to avoid the data thefts, leaks of confidential information and maintaining the interests of parties.
The Silicon Valley Arbitration and Mediation Center (SVAMC) prescribed a set of draft guidelines on August 31, 2023 for the incorporation of AI in arbitral proceedings. It dealt with issues such as a) respect for the due process, b) non- involvement on instances of decision-making, c) due diligence and compliance in the same, d) gatekeeping the confidentiality of parties and processes involved, etc. The users must acquaint and adapt themselves with the applications of the proposed AI tools. A few practical illustrations for the compliant and non-compliant uses of AI were also published. This draft set by inculcating some minor changes were formally issued as guidelines recently on April 30,2024.
The Court of King’s Bench in Manitoba, Canada, while dealing with the employment of AI assistance in preparing materials for court submissions, instructed that the latter must be reflective of the manner in which such assistance was sought. The guidelines were issued keeping in consideration the ever-evolving landscape of AI and hence the underlying difficulty in pin pointing an exact set of regulations for its responsible use.
JAMS, a leading ADR service provider on April 15th, 2024 also introduced artificial intelligence disputes clause and regulations. They have been brought to address the maintenance of data confidentiality involved in the arbitration proceedings. A routine inspection for compliance with the same has also been proposed. They cover AI hardware, software, models as well as training data. Limiting the access to data and materials, they can now be made available only to one or more experts, agreed upon mutually by the parties / Arbitrator/ Tribunal. This feature extends to the “attorney’s eyes only” set up, where the parties mutually decide the materials to be accessible only by the other party attorneys.
Conclusion
The rapid expansion of technology serving several benefits to the humankind is being welcomed around the globe with open arms. The integration of AI into arbitration is an emerging field where new developments and updates keep hitting the news every now and then. A keen eye for the same has to be kept. It will be an interesting watch to see how arbitration in the upcoming years evolves itself with AI as one of the core components in its operation. The various ways in which the technological assistance is being sought in this industry has already been discussed earlier. With the advancements in AI and introduction of new features like AI General Intelligence which imparts it human like abilities, allowing it to self-teach without any input of updated language models; it may or may not act as a factor in reducing human interference. Such speculations can only be answered by time, while we keep ourselves informed of the subsequent evolutions.
As Campbell said in his remarks about AI at ICCA Hongkong’24 that the incline towards AI is due to increasing client needs, “[Clients] desperately need predictability and reliability” and they need it as soon as possible. “If AI is able to resolve disputes with a reasonable level of accuracy, that is where stakeholders are going to go.”
Author(s)
Tishya Vishwanath
Student at RMLNLU, Lucknow