[EDRM Editor’s Note: The opinions and positions expressed are those of Leslie O’Neal and the Hon. Ralph Artigliere (Ret.).]
As artificial intelligence becomes more embedded in our daily lives, disputes involving this complex technology are on the rise. Stakeholders in disputes involving AI are particularly concerned about confidentiality of system processes and algorithms. Disputes involving technology like AI involve experts and complex discovery issues. How do we fairly and efficiently resolve these issues when AI itself complicates process? JAMS has taken a proactive step forward by crafting tailored arbitration rules specifically designed for AI-related disputes. See JAMS Artificial Intelligence Dispute Clause and Rules (effective June 14, 2024). These new rules capitalize on the strengths of arbitration—its speed, flexibility, and efficiency—to keep pace with the rapid evolution of technology.
Arbitration offers
Speed and Efficiency: The arbitration process is typically faster due to less procedural backlog compared to courts.
Confidentiality: Arbitration keeps technical and business-sensitive information private, avoiding public exposure that comes with court cases.
Customizability: Arbitration allows the selection of arbitrators with specific expertise in AI, which courts cannot match with juries and with judges due to random case assignment.
Arbitration is uniquely suited to handling these disputes: parties can modify the process to reduce cost and delay, confidentiality is preserved with less danger of public exposure, and arbitrators with specialized technical expertise can be selected for more informed decision-making. Unlike traditional court systems, where rulemaking is often slow and reactive, JAMS’s approach aims to address challenges before they become systemic obstacles. In the sections that follow, we’ll explore how these new rules are setting the stage for effective, informed dispute resolution in an AI-driven world.
The Growing Need for Tailored AI Rules
Technology is the fastest-growing sector of the global economy, driven by rapid advancements in areas like artificial intelligence and cybersecurity. As AI continues to develop at breakneck speed, it sparks competition, raises ethical questions, and often pushes the boundaries of regulation. This “ragged edge” of innovation inevitably leads to disputes—between competitors, service providers, consumers, and even those claiming injury or exploitation through AI systems. These disputes are not only numerous but are often uniquely complex, requiring nuanced understanding of both technological intricacies and their legal implications.
For courts and ADR professionals, grappling with the evolving landscape of AI is a formidable challenge. Disputes over algorithmic transparency, AI biases, and data privacy are already challenging courts and ADR professionals. These factors are creating a wave of new and complicated issues that challenge our existing legal frameworks. Traditional legal systems can no longer afford to remain passive in the face of technological revolution. Instead of playing catch-up, dispute resolution must anticipate the issues at hand—staying agile and ahead of the curve.
With tailored AI arbitration rules, JAMS demonstrates its leadership and commitment to staying ahead of technological changes, ensuring that arbitration evolves alongside the technologies shaping our world.
Addressing AI’s Unique Challenges through Arbitration
To meet these challenges, JAMS aims to provide adaptability, expertise, and forward-thinking solutions necessary for this evolving landscape. JAMS’s tailored AI arbitration rules are a necessary step to ensure that dispute resolution keeps pace with technology’s transformative impact.
About JAMS (Judicial Arbitration and Mediation Services)
Founded in 1979 by a retired CA judge, JAMS (Judicial Arbitration and Mediation Services) has been a leader in ADR for decades. It has grown to include a global panel of retired judges and attorneys with dispute resolution centers throughout the U.S. and in Toronto and London. It offers a variety of ADR services, including arbitration, mediation, special masters and neutral analysis across all type of industries.
In the sections that follow, we’ll dive deeper into how these new JAMS AI rules differ from existing arbitration frameworks and why these differences matter for those seeking fair, efficient resolutions of AI disputes. These rules are not just theoretical—they’re designed to meet the unique complexities of AI disputes head-on. By examining how these rules differ from standard JAMS and AAA arbitration frameworks, we can understand the real-world impact they’re designed to have on resolving disputes effectively and efficiently.
These new rules capitalize on the strengths of arbitration—its speed, flexibility, and efficiency—to keep pace with the rapid evolution of technology.
Leslie O’Neal and Hon. Ralph Artigliere (Ret.).
As AI continues to evolve, staying ahead demands thoughtful and adaptive regulation—a challenge that JAMS is taking on with these innovative new rules. We invite all stakeholders to engage with additional forward-thinking solutions as we shape the future of AI dispute resolution together.
When Do JAMS Rules Apply?
The JAMS Artificial Intelligence Dispute Clause and Rules (effective June 14, 2024)(hereafter referred to as “the JAMS AI Rules”), apply to binding arbitration if the parties agree to use them, either in a contract or when an AI-related dispute arises. The JAMS AI Rules begin with a Model Dispute Resolution Clause that calls for arbitration under the JAMS Artificial Intelligence Dispute Rules and a comprehensive Model Protective Order. Developed by JAMS Neutrals, Ryan Abbott, M.D., Esq. FCIArb and Daniel B. Garrie, Esq., the rules address some of the challenges AI-related disputes pose. As stated by Abbott and Garrie:
Model Dispute Resolution Clause
The Parties agree that any and all disputes, claims or controversies arising out of or relating to this Agreement shall be submitted to JAMS, or its successor, for mediation, and if the matter is not resolved through mediation, then it shall be submitted to JAMS, or its successor, for final and binding arbitration. Any dispute, controversy or claim arising out of or relating to this Agreement or the breach, termination, enforcement, interpretation or validity thereof, including the determination of the scope or applicability of this agreement to arbitrate, will be referred to and finally determined by arbitration in accordance with the JAMS Artificial Intelligence Dispute Rules. The seat of the arbitration will be [location]. The language to be used in the arbitral proceeding will be English. Judgment upon the award rendered by the Arbitrator(s) may be entered by any court having jurisdiction thereof.
AI systems have the potential to result in massive social benefits and value, but they also involve significant risks. When problems happen—and they will happen—it is important to be able to resolve disputes in a fair, efficient and just manner. Unfortunately, conventional litigation may not achieve that type of outcome and may involve extensive delays, high discovery costs and burdens, and the loss of confidentiality.” Abbott & Garrie, How the JAMS AI Rules Will Improve Dispute Resolution, JAMS ADR Insights (May 3, 2024).
The rules are not intended to address the use of AI in the process of dispute resolution itself (e.g., neutral and advocate use of generative AI systems). However, the Silicon Valley Arbitration and Mediation Center (www.svamc.org) published Guidelines on the Use of Artificial Intelligence in Arbitration on April 30, 2024), covering AI use by all participants in an arbitration. While not binding, these guidelines are useful in dealing with AI use in arbitration proceedings.
Distinguishing JAMS AI Rules from Other Arbitration Rules
The primary differences between the JAMS AI Rules and other commercial arbitration rules are the confidentiality and discovery provisions.
Confidentiality: Unlike other arbitration rules, Rule 26(b) of the JAMS AI Rules requires “all parties and counsel to maintain confidentiality of the proceedings and the award.” In contrast, the JAMS Comprehensive Arbitration Rules require only JAMS and the arbitrator to maintain confidentiality. See JAMS Comprehensive Arbitration Rules & Procedures, Rule 26(a). Unlike the JAMS AI Rules, standard arbitration rules require issuance of a specific order of confidentiality to protect the confidentiality of proprietary information, trade secrets or other sensitive information. See JAMS Comprehensive Arbitration Rules & Procedures, Rule 26(b).
Additionally, the JAMS AI Rules include an “AI Disputes Protective Order” (Appendix A), which automatically applies to protect confidential information. The Protective Order provides that, “All information produced or disclosed in the Action shall be used solely for the prosecution or defense (including any appeal therefrom) of the Action and shall not be used for any other purpose.”
The Protective Order broadly defines “discovery material” and “documents.” It allows information to be designated “Confidential” or “Highly Confidential.” Disclosure of such information is limited. Anyone possessing such information must maintain it in a reasonably secure manner and not disclose the information or any extracts, excerpts, summaries or any testimony, conversations or presentations. Non-parties producing such information may be included in the Protective Order. All experts, advisors, consultants, fact witnesses or potential fact witnesses who receive Confidential or Highly Confidential information must execute a written acknowledgment of the Protective Order.
Automatic protection of confidential information under the JAMS AI Rules is more efficient and cost-effective than the process for obtaining a protective order in arbitration or in state or federal court, which may require filing a motion, a memorandum of law and, in some cases, scheduling and attending a hearing with the judge or magistrate, before a protective order is entered.
Mediation Using the JAMS AI Model Dispute Resolution Clause
The JAMS AI Model Dispute Resolution Clause suggests that disputes should be mediated before proceeding to arbitration. However, the clause lacks specificity regarding the time frame for mediation and does not outline a method for choosing a mediator. It also does not establish mediation as a condition precedent to arbitration, which could result in disputes over the process and potential delays.
Unlike the AI Arbitration Rules, the AI Mediation Clause does not explicitly require a confidentiality agreement or protective order before mediation begins. Given the proprietary nature of AI information, this omission could lead to significant risks, especially since mediation confidentiality varies widely based on the state or court handling the dispute. While the AI Arbitration Rules include a protective order to safeguard sensitive information, it may be beneficial for parties to revise the mediation provisions to extend similar protections.
Although twelve states and the District of Columbia have adopted the Uniform Mediation Act, which offers broad confidentiality protections, other states, such as New York, lack a comprehensive statewide mediation confidentiality statute. Additionally, mediation confidentiality in federal court is inconsistent, as each district court has its own ADR rules, and there is no “mediation privilege” in FRE 501.
Given these considerations, parties using the JAMS AI Model Dispute Resolution Clause in a contract should consider revising the Mediation Clause to include explicit confidentiality requirements and other essential procedural provisions to better protect sensitive information and ensure a smoother dispute resolution process.
Expert and Discovery Provisions: Another major distinction between the JAMS AI Rules and other arbitration rules is the information production process. For example:
- Under JAMS AI Rules, Rule 16.1(b), production and inspection of any AI Systems or related materials (including hardware, software, models and training data) is limited to the Disclosing Party making the materials available to one or more expert(s) in the secured environment the Disclosing Party establishes. The materials shall not be removed from this environment.
- Under JAMS AI Rules, Rule 16.1(b), the Arbitrator shall designate expert(s) to inspect AI systems or related materials at the parties’ joint request. The rule provides that the Arbitrator must use JAMS’ list of third-party experts first in selecting experts; however, since the JAMS expert list has not yet been published, parties must designate experts to perform the inspections.
- To prevent “fishing expeditions” production requests are limited to documents “directly relevant to the matters in dispute” and must be “reasonably restricted” in time frame, subject matter and persons or entities. Broad phraseology and extensive definitions are prohibited, and the Arbitrator may edit or limit the number of requests. See JAMS AI Rules, Rule 16.1(c)(1)-(3).
- There is an expedited schedule for completing discovery (75 calendar days after the preliminary conference for fact discovery and 105 calendar days after the preliminary conference for expert discovery). The hearing is to commence within 60 calendar days after the fact discovery cutoff. The Arbitrator may extend these deadlines (or any procedures in the Rules) for good cause. See JAMS AI Rules, Rule 16.1(h)-(i).
Types of Disputes Involving AI
There are numerous cases arising over the development of generative AI products and platforms, such as chatbots, image synthesis tools, and content creation systems, as well as the use of these products causing harm to users and third parties. Potential parties on either side of the ‘v’ include individuals and entities building AI models, such as developers of open-source or custom models, and those training these models or generating, collecting, or curating the data. It also includes parties integrating models into broader systems, those working with sophisticated hardware, as well as individuals or enterprises using AI systems or licensing platforms.
Problems can arise at any point in this ecosystem, from the initial development stages to the final integration of AI systems into complex environments. For instance, a dispute may occur when an AI developer fails to deliver a model that integrates smoothly with a company’s hardware, resulting in operational disruptions. These complex systems can fail in unpredictable ways, leading to disputes over both technical and legal issues.
Problems can arise at any point in this ecosystem, from the initial development stages to the final integration of AI systems into complex environments.
Leslie O’Neal and Hon. Ralph Artigliere (Ret.).
Cases may include disputes over copyright infringement, such as when generative AI tools unlawfully incorporate content from training datasets without proper authorization. Privacy violations may occur when AI systems collect or process data in ways that contravene regulations such as the EU AI Act or fail to comply with emerging AI-related laws at state and federal levels. The parties involved may range from individual developers to tech startups, healthcare providers using AI diagnostics, multinational corporations employing AI in business operations, or financial institutions leveraging AI for decision-making.
To illustrate these types of disputes more concretely, let’s consider hypothetical scenarios where the JAMS AI rules would play a crucial role in resolving AI-related conflicts.
Hypothetical Examples of Resolving AI Disputes Using the JAMS AI Rules
- Product Liability in AI-Integrated Medical Devices: A healthcare provider utilizes an AI-powered diagnostic tool integrated into a medical device. A patient receives an incorrect diagnosis, resulting in delayed treatment and significant health complications. The healthcare provider sues the AI developer, arguing that the algorithm’s analysis was faulty and caused harm to the patient.
JAMS AI Rules Advantages: In this scenario, JAMS’s specialized arbitration framework allows for technical expert testimony regarding the AI diagnostic tool’s performance. The rules provide for the controlled inspection of the AI system, where experts can review its algorithms and training data within a secure environment without compromising proprietary information. Such an approach not only ensures a precise technical evaluation of the algorithm but also protects the intellectual property of the AI developer, which could be compromised in open court. Moreover, the expedited nature of arbitration is beneficial, as resolving the dispute quickly is in the best interest of the patient and the healthcare provider.
- Contractual Dispute Over AI System Performance: A logistics company licenses an AI system to optimize its supply chain. After implementation, the company claims that the AI underperformed, causing significant delivery delays and financial losses. The AI developer contends that the logistics company did not provide the necessary quality of data for the AI to perform optimally, leading to the system’s perceived failure.
JAMS AI Rules Advantages: The JAMS AI arbitration process allows for an in-depth examination of the contractual obligations of both parties, with an emphasis on whether the conditions for optimal AI performance were met. Arbitration under JAMS rules facilitates the appointment of AI experts who can provide insight into the quality of the data and determine whether the system’s performance issues were due to improper integration by the logistics company. This technical clarity is often difficult to achieve in court, where the specificities of machine learning system requirements might be lost on generalist judges or juries.
The confidentiality provisions in the JAMS AI rules offer a significant advantage, as they ensure that proprietary information about the AI system and its integration process remains protected. This prevents sensitive operational details from being publicly disclosed, which could be a major concern for both the logistics company and the AI developer. Additionally, the arbitration setting fosters a cooperative resolution, focusing on finding a solution that allows both parties to continue their business relationship.
Both of these examples demonstrate that disputes involving artificial intelligence—whether related to intellectual property, medical devices, or contractual performance—can benefit significantly from the structure and expertise that JAMS AI-specific arbitration rules provide. But those lessons apply in many AI-involved cases. By leveraging expert-led reviews, secure and confidential processes, and tailored procedural flexibility, arbitration through JAMS offers an effective and efficient way to resolve complex issues in the AI domain, ensuring that all stakeholders achieve fair and timely outcomes. As AI continues to shape industries, tailored dispute resolution mechanisms like those offered by JAMS will be key to mitigating risks while supporting innovation.
CONCLUSION
The explosion of AI in a competitive marketplace of product developers, users, and impacted third parties is bound to create complex and multifaceted cases in need of efficient and just resolution. The need for confidentiality and efficient exchange of information makes arbitration a better way to resolve AI disputes. Fortunately, JAMS has tailored its rules to tackle some of the most difficult issues. These specialized rules can result in fairer, faster and more cost-effective outcomes.
By tailoring its rules to address the complexities inherent in AI disputes, JAMS is providing dispute resolution mechanisms not only fit for today but also adaptable for the future. As AI continues to revolutionize industries and present new challenges, this approach will benefit current stakeholders and set a precedent for how the legal community adapts to technological advancements.
Assisted by GAI and LLM Technologies per EDRM GAI and LLM Policy.
