AI Liability and EU Harmonized Standards

This chapter examines AI Liability in the EU and focuses on how it operates at the intersection of ex-ante and ex-post mechanisms, combining the EU AI Act&s;s pre-market obligations with the post-market accountability established by the Product Liability Directive (PLD), adopted in 2024. Central to this dual regime is the role of harmonised safety standards, which serve both as benchmarks for compliance and as reference points in determining liability. However, the precise legal weight of such standards remains unclear, particularly in the context of AI systems. To date, case law directly addressing how safety standards influence liability for AI-related harm is very limited. The article argues that a clearer understanding of the function of safety standards is essential to ensuring coherence between ex-ante regulatory compliance and ex-post liability and that a robust ex-post liability regime remains critical to safeguarding individuals and ensuring trust in AI systems..