The AI Liability Puzzle: Rethinking Defective Product Liability for AI

The impetus of AI is revealing numerous, incredibly diverse, and cross-sector applications. The benefits associated with the use of AI and expected from their systematic and extensive application are multiple, extremely promising, and to a certain extent overwhelmingly positive. Nevertheless, the expansive and growing use of AI in our society can also be a source of new risks, lead to undesired outcomes and unintended consequences, or raise legal concerns and social challenges of many different kinds. In the face of such potentially negative effects, the fundamental question is whether traditional legal regimes are equipped to manage the risks and effectively resolve the conflicts arising from these situations in complex technological environments. The adequacy and completeness of civil liability regimes in the face of technological challenges have an extraordinary societal relevance. Should the liability system reveal insufficiencies, flaws and gaps in dealing with damages caused by AI, in particular, victims can remain uncompensated or, at least, only partly compensated. The social impact of a potential inadequacy of existing legal regimes to address new risks created by AI might then compromise the expected benefits. This Chapter discusses first the inadequacies detected in the existing civil liability regimes in the face of AI to explore the different policy options to consider with the aim of accommodating the liability system to scenarios of damages caused by, or with the intervention of, AI systems. Subsequently, the Chapter analyses the European Union’s response to the AI liability challenges with the adoption of two legislative proposals to accommodate product liability rules as well as some civil liability ones to damages caused by, or with the intervention of, AI systems (the latter was finally withdrawn in 2025). The Chapter focuses on the finally adopted revision of the Product Liability Directive, identifying the main challenges, and exploring the primary solutions, and elaborates on the idea that the revised rules play a key role in the policy strategy to resolve the AI Liability puzzle.