The European Parliament has approved the AI Liability Directive, establishing a legal framework that holds developers and deployers of AI systems financially responsible for damages caused by their products. The directive, effective January 1, 2027, creates a presumption of causality when AI systems classified as high-risk cause harm, shifting the burden of proof from victims to companies.
Under the new rules, individuals harmed by AI-driven decisions in healthcare, employment, financial services, or law enforcement can seek compensation without needing to understand the technical workings of the system. Courts can order companies to disclose technical documentation and training data methodologies.
The directive is expected to influence similar legislative efforts in the United States, Canada, and Australia, as multinational companies may adopt EU compliance standards globally.
