G20 Ministerial Statement on Trade and Digital Economy (Section I, Subsection 3 "Human-centered Artificial Intelligence")

• Authority: Other multinational
• Jurisdiction: Multinational
• Collections: Multinational
• Most recent activity: Proposed
• Most recent activity date: 2019-06-09

Risk subdomain coverage

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

• Unfair discrimination and misrepresentation
• Unequal performance across groups
• Compromise of privacy
• AI system security vulnerabilities
• Overreliance and unsafe use
• Loss of human agency and autonomy
• Power centralization and unfair distribution
• Increased inequality and decline in employment
• Governance failure
• Environmental harm
• AI pursuing its own goals
• 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, with brief mentions of privacy (2.1), security (2.2), discrimination (1.1), labor market transitions (6.2), and governance (6.5). Coverage is primarily at the principle level rather than detailed risk mitigation measures. Most subdomains receive coverage scores of 1-2, with no subdomain receiving a score of 3.

Sector coverage

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

• Information
• Scientific Research & Development Services
• Educational Services
• Public Administration (excl. National Security)

All other sectors had no coverage in this document.

Sectors Summary

This is a cross-sectoral governance document that does not target specific economic sectors. It provides general principles for human-centered AI applicable across all sectors, with emphasis on enabling innovation and addressing broad societal challenges.

Legislative Status

Soft Law

Legislative Status Summary

This is a non-binding ministerial statement that uses voluntary language and explicitly states the principles are 'non-binding'. It relies on voluntary cooperation and commitment rather than legal enforcement.

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Entities: G20; OECD
• Summary: The G20 ministerial statement draws from OECD Recommendation on AI, indicating both G20 and OECD as proposers of the governance principles.

Target

• Entity Types: AI_Governance_Actor, AI_Developer, AI_Deployer
• Entities: G20 members; governments; international organizations; academia; civil society; private sector; MSMEs; startups
• Summary: The document targets all stakeholders including governments, private sector, academia, and civil society, with particular focus on MSMEs and startups adopting AI.

Enforcer

• Summary: No enforcement mechanisms or enforcing bodies are specified in this non-binding statement.

Monitor

• Entity Types: AI_Governance_Actor
• Entities: G20 Repository of Digital Policies
• Summary: The G20 Repository of Digital Policies is mentioned as a mechanism for information sharing, suggesting a monitoring/tracking role.

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 does not focus on specific lifecycle stages but rather provides high-level principles applicable across the entire AI lifecycle. It emphasizes research and development, adoption, and responsible use without detailed stage-specific requirements.

Technical Scope

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

No technical scope terms identified.

Technical Scope Summary

The document uses the general term 'AI' and 'AI technologies' without distinguishing between models, systems, or specific types of AI. It does not mention frontier AI, general purpose AI, foundation models, generative AI, or compute thresholds.