AI and Arbitration: A perspective from France

France has adopted a measured and ethical approach to the integration of AI in arbitration. The jurisdiction has embraced AI as a tool for efficiency, while simultaneously reinforcing strict legal, procedural, and ethical boundaries. This article explores how AI is shaping the arbitration workflow today in France, how regulators and courts are responding to its integration, and what lies ahead for a legal system known for balancing innovation and legal tradition.

AI’s Role in Arbitration and the French Legal Landscape

AI is increasingly built into the arbitration process in France. Whether a dispute is straightforward or complex, AI tools have proven valuable in simplifying dense legal documents, facilitating research, and mapping out intricate legal precedents. French practitioners now routinely use advanced legal tech platforms such as Predictice and Lexis+ AI for legal research and analysis, Hyperlex AI for contract management, Luminance and Relativity for summarizing legal documents, LinnkAI for translating multilingual legal documents, and Lami Liaisons Solutions for risk and compliance management.

Digital Reformation under the French Arbitration Law

France’s arbitration framework is among the most forward-looking in Europe, and it continues to evolve.[1] A notable milestone was the release of the “Report and Reform Proposals” by the Working Group on the Reform of French Arbitration Law, co-chaired by François Ancel and Thomas Clay.^[2] Although the report does not directly regulate AI, it does discuss the importance of digitalisation in arbitral proceedings, e.g. by proposing formal recognition of electronic awards, potentially paving the way for AI-assisted drafting, provided human oversight is preserved.

Discussions around AI also took centre stage at the Paris Arbitration Week (PAW) this year where panels such as “AI in Arbitration: Less Talk, More Tech” and “The Use of AI in Arbitration” identified practical applications – computer programs that can sort through thousands of pages in seconds, assist with case strategy, and even simulate procedural scenarios. The Paris International Arbitration Chamber (CAIP) has undertaken a commendable endeavour to provide a useful bilingual guidance on the use of AI in arbitrations under its rules. Additionally, the Paris Bar has partnered with Ordalie, an AI model to assist law firms in their regular operational tasks whilst encouraging effective time management.

Synergy Between French and EU Law

The adoptionof the EU AI Act is reshaping the AI regulatory landscape across Europe.[3] Rather than drafting its own standalone AI legislation, France is working within the European framework while developing sector-specific guidelines.

For instance, proposed amendments to the French Intellectual Property Code (IPC) aim to regulate the use of generative AI and ensure AI-generated outputs respect copyright norms.[4] Additionally, the French data protection agency (CNIL) has published a four-part AI Action Plan[5] focused on promoting innovation while safeguarding data privacy and accountability.

Under the EU AI Act, AI tools used in arbitration for research, legal analysis or fact interpretation are classified as high-risk, but those used for administrative tasks are practically excluded from that category.^[6] This aligns with French arbitration law, Article 1450 of the Code of Civil Procedure, which mandates that arbitrators be natural persons. Together, these legal frameworks reinforce a consistent principle that AI is going to remain in a completely supportive role, although it may increase efficiency in case management, document review, and predictive analysis. However, it should not replace the human element in decision making, at least not in the near future.

AI Before French Courts: Lessons for Arbitration

The French courts have been a laboratory for figuring out how much AI can contribute to law and where its red lines must be drawn. On 28 April 2025, the Working Group on Artificial Intelligence submitted its report titled, “Cour de Cassation and Artificial Intelligence: Preparing the Court of Tomorrow”,^[7] highlighting both the opportunities and risks of AI. The report confirms that there is currently no need for AI in decision-making roles, reinforcing the principle that adjudication must remain human-led.

Furthermore, the Cour de Cassation has also collaborated with AI research teams such as the ALMAnaCH project of INRIA to detect inconsistencies in case law.^[8] The court stressed these systems – internally developed, GDPR-compliant, and subject to strong governance – assist legal clerks in analysis without displacing human reasoning.

Meanwhile, a ruling by the Tribunal Judiciaire de Nanterre in February 2025 cautioned against deploying AI surveillance tools in the workplace without prior consultation with employee representatives.^[9] This reaffirms the duty of transparency and stakeholder disclosure – principles that also apply in arbitration when AI tools are used in case preparation.

At a November 2024 colloquium hosted by the Tribunal Administratif d'Orléans, judicial experts examined AI’s role in high-tech fields such as construction, finance, and environmental law.^[10] While the specialists concurred that AI has the capacity to speed up data processing and pattern detection, the debates unleashed eternal threats: algorithmic prejudice, lack of transparency in “black box” systems, and risk of dissolving impartiality. The consensus was clear: AI may assist, but its results should be verifiable, disputable, and subject to human judgment.

While these decisions are not specifically talking about arbitration, they underscore values that are fundamental to arbitration: due process, transparency, and party autonomy.

Looking Ahead: Reform, Resistance, and a Human-Centric Future

According to the 2025 International Arbitration Survey by Queen Mary University of London and White & Case, 91% of respondents expect to use AI for research and data analytics within the next five years. Respondents largely approve of the use of AI by arbitrators to assist in administrative and procedural tasks. There is strong resistance, however, to its use for tasks requiring the exercise of discretion and judgment, which are fundamental aspects of the mandate given to arbitrators.^[11]

To conclude, the message from the French court is consistent: AI may accelerate the arbitration process, but it must not substitute a human’s judgment. The message is as expressed at a PAW event last year, “[t]echnology can accelerate the ride, but the driver's seat should still be human”.[12]

This article was originally published on Daily Jus on Friday 1st of August, with thanks to Jus Mundi & Jus Connect.