The Future of Civil Liability for Artificial Intelligence Systems after the Withdrawal of the European AI Liability Directive

Abstract

The European Union proposed an Artificial Intelligence Liability Directive in September 2022 with the aim of modernising the legal framework and ensuring a balance between the protection of victims’ rights and the promotion of technological innovation. However, several factors led to the withdrawal of the proposal in February 2025, including political pressures, the influence of major corporate lobbying groups, the adoption of Regulation (EU) 2024/1689 establishing a risk-based regulatory framework for AI systems, and the revision of the Product Liability Directive concerning liability for defective products.

The revised Product Liability Directive facilitates compensation claims for damage caused by defective products and eases the burden of proof on victims. The withdrawn AI Liability Directive was based on existing European case law but was adapted to address the specific challenges posed by artificial intelligence. Consequently, legal principles developed through judicial decisions were transformed into clear legislative rules tailored to the particularities of AI systems.

The withdrawal of the Directive does not mean the disappearance of the underlying legal principles. National courts dealing with AI-related disputes will continue to apply the principles of European Union law, particularly the principle of effectiveness. They may rely on general rules of liability and established legal principles to apply presumptions of causation in a flexible manner that ensures effective protection for individuals harmed by AI systems. They may also draw upon the case law of the Court of Justice of the European Union, particularly in matters relating to evidence. The future legal framework therefore requires a comprehensive approach that promotes sustainable innovation while ensuring adequate protection for victims.