AI Liability Directive

• Authority: European Union
• Jurisdiction: Multinational
• Collections: Multinational
• Most recent activity: Proposed
• Most recent activity date: 2022-03-03

Risk subdomain coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the risk subdomain are explicitly described.

• Governance failure

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.

• AI system security vulnerabilities
• Lack of capability or robustness
• Lack of transparency or interpretability

All other risk subdomains had no coverage in this document.

Risk Coverage Summary

The document has minimal coverage of risk domains, primarily addressing governance failure (6.5) and AI system safety failures (7.3, 7.4). The directive focuses on legal liability mechanisms rather than substantive AI risks, with brief mentions of security vulnerabilities (2.2) and lack of robustness (7.3) in the context of establishing fault conditions.

Sector coverage

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the sector are described.

• Trade, Transportation & Utilities
• Finance & Insurance

All other sectors had no coverage in this document.

Sectors Summary

This directive does not govern specific economic sectors but rather establishes horizontal liability rules applicable to high-risk AI systems across all sectors. The only sector explicitly excluded is transport, where existing EU liability rules apply.

Legislative Status

Hard Law

Legislative Status Summary

This is a binding EU Directive with mandatory obligations on Member States to transpose into national law, featuring enforcement mechanisms, mandatory language throughout, and legal obligations for courts and parties.

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Entities: European Union; European Parliament; Council of the European Union
• Summary: The document is an EU Directive proposed and adopted by the European Parliament and Council, as indicated in the closing articles and the legislative structure.

Target

• Entity Types: AI_Developer, AI_Deployer
• Entities: Providers of high-risk AI systems; Users of high-risk AI systems; Persons subject to provider obligations pursuant to Article 24 or Article 28(1) of the AI Act
• Summary: The directive targets providers and users of high-risk AI systems who may be defendants in civil liability claims, as well as persons who assume provider obligations under the AI Act.

Enforcer

• Entity Types: AI_Governance_Actor
• Entities: National courts of Member States; European Commission
• Summary: National courts are empowered to order evidence disclosure, apply presumptions, and adjudicate liability claims. The Commission monitors implementation and reviews the directive's application.

Monitor

• Entity Types: AI_Governance_Actor
• Entities: European Commission; Member States
• Summary: The Commission establishes a monitoring program and reviews the directive's application. Member States must communicate relevant data and evidence to the Commission for monitoring purposes.

AI Lifecycle Stages

Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems

• Plan and Design
• Collect and Process Data
• Build and Use Model
• Verify and Validate
• Deploy
• Operate and Monitor

AI Lifecycle Summary

The document primarily addresses the Deploy and Operate and Monitor stages of the AI lifecycle, with implicit coverage of Build and Use Model and Verify and Validate stages through references to AI Act requirements on training data, design, development, accuracy, robustness, and cybersecurity.

Technical Scope

Terms relating to technical scope that are mentioned or implicitly referenced in the document

• AI Systems

Technical Scope Summary

The document explicitly mentions AI systems and high-risk AI systems as defined in the EU AI Act. It does not explicitly mention frontier AI, general purpose AI, task-specific AI, foundation models, generative AI, predictive AI, open-weight models, or compute thresholds.