Liability for AI-related products in the EU—a state of play
EU Law analysis: The evolving landscape of artificial intelligence (AI) creates the need for a robust legal framework to address the unique challenges posed by AI-related products. This commentary explores three key areas: the current regime under the EU Product Liability Directive (PLD), the now abandoned AI Liability Directive (AILD), and the Revised EU PLD set to take effect on 9 December 2026. The current PLD framework has struggled to address AI-induced damages, highlighting the need for the current rules to be updated. The AILD proposal aimed to incorporate new rules that would help to bridge the gap between damage caused by AI-related products and more traditional products. These new rules had included a proposal to change the burden of proof rules by introducing a presumption of causality. These changes would have offered harmonisation and a lower burden on victims claiming compensation for damages for AI related products. Following the withdrawal of the AILD these gaps remain. The Revised EU PLD will expand to include AI, addressing defectiveness, responsibility within the value chain, and substantial modifications. Practical examples illustrate the new regime's application, while challenges for victims and impacts on the AI sector highlight the importance of risk mitigation strategies. Written by Susan Walsh, consultant, and Aoife Keenan, associate at William Fry.