Policy Briefing — EU AI Liability Directive Proposal
The European Commission proposed the AI Liability Directive to ease the burden of proof for victims seeking compensation from AI-related harm.
The draft directive introduces rebuttable presumptions of causality when AI deployers fail to provide evidence or comply with transparency obligations. It harmonises disclosure requirements so claimants can request information about high-risk AI systems.
- Burden of proof. Courts can presume causality if an AI system does not meet AI Act obligations and harms occur.
- Disclosure rights. Claimants may obtain evidence about high-risk AI, subject to confidentiality safeguards.
- Alignment with AI Act. The directive complements the AI Act by adding remedies for damages when obligations are breached.
Risk managers should update EU liability assessments to account for the proposed presumptions and documentation duties.