The Role of Artificial Intelligence in Arbitration Proceedings

Introduction

The rapid advancement of artificial intelligence (“AI”) has profoundly transformed everyday human activities, and the field of arbitration is no exception. McKinsey research sizes the long-term AI opportunity at $4.4 trillion in added productivity growth potential from corporate use cases.[1] In an effort to embrace this transformative technology, arbitration proceedings are increasingly leveraging AI to streamline processes, accelerate dispute resolution, and enhance overall efficiency. However, this pursuit of efficiency raises a critical question: what are the implications when technological precision begins to eclipse human judgment?

AI should not replace the arbitrator’s role in decision-making or award drafting; rather, it should serve as a complementary tool to enhance, not supplant, human expertise. Recognizing both the promise and the perils of AI integration, several arbitral institutions are taking proactive steps to establish ethical frameworks for its use. Notable among these are the Chartered Institute of Arbitrators (“Ciarb”), which released its 2025 Guidelines on the Use of Artificial Intelligence in Arbitration, and the Silicon Valley Arbitration & Mediation Center (“SVAMC”), which has published its own set of AI guidelines. This article explores the evolving role of AI in arbitration.

Advantages of AI integration in Arbitration

One of the most significant contributions of AI to international arbitration is the improvement of efficiency in legal research and document review. Traditionally, this task required arbitrators or lawyers to manually read and analyze large volumes of documents, a process that was time-consuming and labour-intensive. AI has transformed this approach. With the help of machine learning and natural language processing, systems like IBM's Watson, Arbitrators can quickly sift through large datasets, identify relevant information, and analyze submissions from the opposing party. This leads to faster decision-making, reduced costs, and makes arbitration more accessible and practical for parties involved.[2]

Speech recognition technology is another AI-powered innovation that is increasingly being used in international arbitration. This technology can accurately transcribe spoken words, distinguish between different speakers, and translate languages when needed. These features are especially valuable in arbitration, where real-time transcription and translation are often essential. By reducing the need for human interpreters and speeding up transcription, AI helps lower costs and improve the efficiency of hearings. Furthermore, AI’s speech recognition can assist in drafting arbitration awards by automating standard sections, allowing arbitrators to focus more on complex aspects of the case. This not only saves time but also promotes greater accuracy and consistency in the drafting process.[3]

Current international legal framework for AI in Arbitration

The increasing integration of Artificial Intelligence (AI) into legal and dispute resolution processes has sparked global discussions on the need for regulation and ethical oversight. AI algorithms, while powerful, can reproduce and magnify the biases contained in the data on which they are trained. This means that in arbitration, where fairness, impartiality, and sound judgment are critical, AI systems could influence outcomes or pattern recognition in ways that undermine these values. In response to such concerns, several legal and regulatory frameworks have emerged. The most notable to date is the European Union’s Artificial Intelligence Act, adopted in 2024. This legislation is the first of its kind globally and represents a comprehensive attempt to regulate AI technologies by managing their risks and fostering innovation responsibly.[4]

Furthermore, the core objective of the AI Act is to ensure that AI systems used within the EU are safe, uphold the fundamental rights of individuals, and encourage responsible technological advancement. The regulation classifies AI systems into four categories based on their associated risks: minimal risk, limited risk, high risk, and unacceptable risk. High-risk AI systems which include those deployed in sensitive sectors such as education, healthcare, banking, or law enforcement are subject to strict requirements. These include obligations for transparency, accuracy, traceability, and human oversight. On the other hand, unacceptable-risk systems are outright banned. Examples of such prohibited use include facial recognition in public spaces without consent, biometric categorization based on sensitive personal attributes (e.g., race, gender, or political opinions), and AI systems that exploit user vulnerabilities or manipulate behavior.[5]

The regulation also mandates that users be informed when content is generated or modified by AI. Furthermore, non-compliance with the law may attract substantial penalties, which underscores the seriousness of the regulatory expectations. Notably, the regulation has an extraterritorial scope, i.e. it applies not just to EU-based companies but also to non-EU entities whose AI systems are used within EU territories. This positions the EU Act as a global benchmark for AI governance and responsible technology use.[6]

Complementing broader legal efforts like the EU AI Act, specialized arbitration bodies have begun creating tailored guidelines to manage AI's integration into arbitral procedures. A leading initiative in this area came from the Silicon Valley Arbitration and Mediation Center (SVAMC), which released its Guidelines on the Use of Artificial Intelligence in Arbitration in 2024. These guidelines are designed to ensure that AI technologies can enhance arbitration efficiency and accessibility without compromising legal integrity or the rights of the parties. They offer both ethical direction and practical advice for using AI at different stages of arbitration. Some key Provisions of the SVAMC Guidelines: Responsible selection and ongoing validation of AI tools, Arbitrator’s non-delegable mandate, Accountability and human oversight; Ethics, Transparency, and Data Security.

Another recent development in the field came from the Arbitration Institute of the Stockholm Chamber of Commerce (SCC), which issued its own guidelines on October 16, 2024. These SCC guidelines aim to enhance arbitration efficiency and reduce the escalating costs and delays that have drawn criticism across global jurisdictions. Unlike the EU AI Act or the binding nature of some legal standards, the SCC’s guidelines are designed to be flexible and adaptable. They provide direction without imposing mandatory rules. Their application is limited to situations where they do not conflict with party agreements, local arbitration laws, or mandatory public policy. The SCC guidelines reaffirm the growing consensus that while AI holds immense potential for streamlining arbitration, its use must be governed by principles that preserve the fairness, impartiality, and human-centric nature of legal decision-making.[7]

Case Study: LaPaglia v. Valve Corp.

In LaPaglia v. Valve Corp (3:25-cv-00833), a consumer, Mr. LaPaglia (the “Claimant”), initiated arbitration through the American Arbitration Association against Valve Corp. (the “Respondent”), the owner of the Steam online gaming platform. He sought damages for alleged antitrust violations that inflated game prices and for breach of warranty over a defective game. The dispute was heard over ten days in December 2024 by a sole arbitrator. According to the Claimant, during breaks, the Arbitrator expressed a desire to issue the award swiftly due to an upcoming trip to the Galapagos. The final briefs were submitted on 23 December 2024, and the Arbitrator issued a 29-page award on 7 January 2025, reportedly the day before his trip.

On 8 April 2025, the Claimant petitioned the U.S. District Court for the Southern District of California to vacate the award on statutory authorities. He alleged that the Arbitrator improperly used Artificial Intelligence (AI) to draft the award, exceeding his authority under the arbitration agreement. The Petition was based on several claims: the Arbitrator had previously stated he used ChatGPT to draft an article to save time, expressed urgency to conclude the case before his trip, and produced an award allegedly containing inaccuracies and AI-like phrasing. The Claimant’s legal team even queried ChatGPT, which identified the language as likely AI-generated.

The Claimant argued that using AI to generate the award violated the arbitration agreement requiring a reasoned decision from a neutral human arbitrator. He likened this situation to precedents where courts vacated awards due to arbitrators misrepresenting qualifications or delegating decision-making. The court has not yet ruled, but the case raises pressing questions about AI’s role in arbitration and the boundaries of arbitrator discretion.[8]

Conclusion

Artificial Intelligence is rapidly becoming an indispensable tool in arbitration, offering remarkable improvements in efficiency, speed, and cost-effectiveness particularly in legal research, document review, transcription, and even award drafting. As demonstrated, AI can significantly enhance the arbitration process by supplementing human efforts and streamlining complex tasks. However, this integration must be approached with caution and guided by clear ethical and legal standards. The ongoing evolution of regulatory frameworks, such as the EU’s Artificial Intelligence Act and specialized guidelines from institutions like SVAMC and SCC, underscores a global recognition of both the opportunities and risks associated with AI in arbitration.

[1]  Hannah Mayer, Lareina Yee, Michael Chui and Roger Roberts, ‘Superagency in the workplace: Empowering people to unlock AI’s full potential’ (McKingsey & Company, 28 January 2025). <https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/superagency-in-the-workplace-empowering-people-to-unlock-ais-full-potential-at-work> accessed 15 June 2025.

[2] Aksa Bethel Linson, ‘The Integration of AI in International Arbitration: Streamlining Efficiency or Introducing New Complexities?’ (The Jurist, 11 October 2025). <https://www.the-jurist.com/article/the-integration-of-ai-in-international-arbitration-streamlining-efficiency-or-introducing-new-complexities> accessed 17 June 2025.

[3] Ibid.

[4] Germán Darío Flórez Acero, ‘The Use of Artificial Intelligence in Arbitration: Friends with Benefits’ ( 30 April 2025). <https://revistas.javeriana.edu.co/files-articulos/VJ/74(2025)/6722763004/index.html> accessed 17 June, 2025.

[5] Ibid.

[6] Ibid.

[7] Ibid.

[8] ‘When Arbitrators Use AI: LaPaglia v. Valve and the Boundaries of Adjudication’ ( Aceris law, 19 April 2025). <https://www.acerislaw.com/when-arbitrators-use-ai-lapaglia-v-valve-and-the-boundaries-of-adjudication/> accessed 20 June 2025.