Ensures AI activities align with human rights, democracy, and rule of law. Requires measures for transparency, accountability, non-discrimination, and privacy protection. Establishes oversight, remedies, and risk management frameworks. Promotes international cooperation and public consultation. Excludes national security and defense activities.
Analysis summaries, actor details, and coverage mappings were LLM-classified and may contain errors.
This is a binding international treaty (Framework Convention) with mandatory obligations on Parties, enforcement mechanisms, oversight bodies, and formal ratification procedures. The document uses mandatory language throughout and establishes legally binding commitments.
The document has good coverage of approximately 10-12 subdomains, with strong focus on discrimination (1.1, 1.3), privacy compromise (2.1), security vulnerabilities (2.2), misinformation (3.1, 3.2), malicious actors (4.1), human-computer interaction (5.1, 5.2), governance failure (6.5), and system safety (7.3, 7.4). Coverage is concentrated in human rights protection, democratic integrity, and AI system reliability domains.
This is a horizontal, cross-sectoral Framework Convention that applies to AI systems across all economic sectors. It does not target specific industries but rather establishes general obligations for public authorities and private actors across all sectors where AI systems may impact human rights, democracy, and rule of law.
The document explicitly covers all AI lifecycle stages from planning through operational monitoring. It repeatedly references 'activities within the lifecycle of artificial intelligence systems' and provides specific requirements for each stage including design, data processing, development, testing, deployment, and ongoing monitoring.
The document explicitly defines and covers AI systems broadly. It does not use terms like 'frontier AI', 'general purpose AI', 'foundation models', or 'generative AI'. It does not specify compute thresholds or distinguish between open-weight and closed models. The focus is on AI systems generally, defined by their functional characteristics.
Council of Europe, Committee of Ministers of the Council of Europe
The document is a Framework Convention developed by the Council of Europe, as indicated by the title and signature location (Vilnius, 5.IX.2024). The Committee of Ministers has authority over amendments and accession decisions.
National oversight mechanisms designated by each Party, Conference of the Parties, Committee of Ministers of the Council of Europe, Secretary General of the Council of Europe
Each Party must establish or designate national oversight mechanisms with enforcement powers. The Conference of the Parties and Committee of Ministers provide international oversight and coordination.
Conference of the Parties, national oversight mechanisms, competent authorities designated by each Party, existing domestic human rights structures
The Conference of the Parties monitors implementation through periodic reporting. National oversight mechanisms and competent authorities monitor compliance at the domestic level, with coordination with human rights structures.
Public authorities, private actors acting on behalf of public authorities, private actors undertaking AI lifecycle activities, member States of the Council of Europe, non-member States, European Union
The Convention applies to both public authorities and private actors engaged in AI system lifecycle activities. It targets Parties (States and the EU) who must regulate these actors within their jurisdictions.
14 subdomains (9 Good, 5 Minimal)