Evolving AI and arbitration legal practices

The impact of artificial intelligence (AI) on society is evident, yet the extent of its influence remains unclear. What is clear, however, is that AI-driven disputes will only increase in the coming years. These disputes will involve a wide range of conflicts from data privacy breaches, intellectual property infringement, unauthorized synthetic content, and trade secret misappropriation involving large language models (LLMs), to breach of contract. As it stands, alternative dispute resolution (ADR) today is largely ill-equipped to handle these types of disputes; in recognition of this, I, along with Dr. Ryan Abbott, worked closely with JAMS to develop JAMS Artificial Intelligence Dispute Resolution Rules (AI Rules). "JAMS Rules Governing Disputes Involving Artificial Intelligence Systems, effective April 15, 2024," www.jamsadr.com/rules-clauses/artificial-intelligence-disputes-clause-and-rules (last visited May 9, 2024).
At a high level, the AI Rules establish that, unless otherwise agreed by the parties, JAMS will propose arbitrator candidates with AI knowledge (Rule 15(b)), mandate a protective order by default to secure sensitive information and stringent data handling during disputes (Rule 16.1(a)), and limit expert testimony to written reports and directed oral responses to maintain focus and confidentiality throughout the arbitration process (Rule 16.1(b)). The net effect of the AI Rules is that they streamline the resolution process and reduce the time and cost associated with resolving AI disputes.

ARBITRATOR SELECTION

Rule 15(b) of the AI Rules provides that "JAMS shall propose, subject to availability, only panelists approved by JAMS for evaluating disputes involving technical subject matter with appropriate background and experience. JAMS shall also provide each party with a brief description of the background and experience of each Arbitrator candidate." Id. at 15(b). This prerequisite helps to alleviate parties expending substantial resources educating the arbitrator on the technical aspects underpinning the dispute and that the arbitrator will be capable of adjudicating appropriately. For example, in a dispute concerning the execution of a funding agreement, the main issue is the alleged misrepresentation of a company's machine learning algorithms' capabilities and performance metrics, which were crucial in securing the funding. The Arbitrator should possess knowledge of AI to adequately understand the technical nuances of the case.

THE DEFAULT PROCEDURE TO ADDRESS CONFIDENTIALITY

In addition, the AI Rules establish a default procedure to address the confidentiality of AI systems in a dispute via Rule 16.1(a). This rule requires the arbitrator to issue an "AI Disputes Protective Order" unless otherwise agreed upon by the parties. Id. at 16 and 26. For example, consider a scenario in which one company accuses another of illegally reverse-engineering its proprietary speech recognition software based on a next-generation AI platform. Here, as part of the JAMS arbitration process, the arbitrator issues a default protective order at the onset of the matter to manage the exchange of sensitive data, including trade secrets and internal testing protocols. Immediately issuing this default order secures that the algorithm, training data, LLM and other sensitive data artifacts will be kept confidential.

THE PRODUCTION AND INSPECTION OF AI SYSTEMS IN A SECURED ENVIRONMENT

The AI Rules also provide specific procedures for producing and inspecting AI systems during discovery. Under this rule, the opposing party does not review AI systems and related materials directly. Instead, a designated expert, agreed upon by the parties or appointed by the Arbitrator, reviews these materials in a secure and confidential environment. Id. at 16.1(b). The expert is prohibited from removing materials or information from this designated environment. For example, in a dispute over the alleged theft of proprietary machine learning models, the AI Rules dictate that direct inspection of AI systems by the opposing party is prohibited. Instead, if the parties agree the Arbitrator may appoint a respected AI expert to review the relevant materials. The expert conducts their examination in a secure environment, ensuring confidentiality. They have controlled access to the AI systems and datasets but cannot remove any materials from this environment. Their findings are then documented and submitted to the arbitration panel to aid in resolving the dispute without exposing sensitive proprietary information. As the example demonstrates, the AI Rules will save the parties both money and time as well as minimize the risk of sensitive information being leaked during the arbitration process.

ARBITRATORS SET EXPERT'S SCOPE OF WORK FOR TECHNICAL ISSUES

In addition, the AI Rules, via Rule 16.1(b), limit expert testimony on technical issues related to AI systems to a written report addressing questions posed by the arbitrator that can be supplemented by testimony during the hearing. For example, in a dispute over the performance of an AI-driven logistics system, complexities arise due to the system's advanced algorithmic decisions. To resolve these technical issues, if the parties agree, the Arbitrator may designate an AI expert The expert prepares a detailed written report addressing specific questions posed by the arbitrator concerning the AI system's design, implementation and operational outputs. Later, the expert is also called to testify, allowing both parties to question their findings and the assumptions underlying their expert analysis. This process makes certain that the technical aspects are thoroughly examined and understood, adhering strictly to the constraints set by the arbitrator.
The JAMS Artificial Intelligence Dispute Resolution Rules are a crucial update of arbitration processes for modern technology. For legal professionals, it is vital to know and use the AI Rules in AI-related contracts and disputes to keep up with the challenges of technology.
Disclaimer: The content is intended for general informational purposes only and should not be construed as legal advice. If you require legal or professional advice, please contact an attorney.
Ryan Abbott, M.D., Esq., FCIArb is a co-creator of the JAMS AI Disputes Clause and Rules and a neutral with JAMS in New York, Los Angeles, and London. Daniel B. Garrie is a co-creator of the JAMS AI Disputes Clause and Rules and a neutral with JAMS, with a focus on cybersecurity, data privacy, e-discovery, and intellectual property.