Recommendation of the Council on Artificial Intelligence

Summary

The first intergovernmental standard on AI, adopted by the OECD Council at Ministerial level: five values-based principles for trustworthy AI and five recommendations to Adherents for national policies and international co-operation. Principle 1.4 says AI systems should be robust, secure and safe throughout their lifecycle, in normal use, foreseeable use or misuse, with mechanisms to override, repair or safely decommission systems that risk undue harm. Other principles call for human agency and oversight, transparency, traceability and a systematic risk-management approach across the AI system lifecycle.

Safety topics

Status notes

Adopted on 22 May 2019 (C(2019)34). On 8 November 2023 the Council revised the definition of 'AI system'; on 3 May 2024 the Council meeting at Ministerial level revised the Recommendation to address misinformation and information integrity in the context of generative AI, uses outside the intended purpose and misuse, and the safe override and decommissioning of AI systems. The Compendium lists no later amendment. The G20 AI Principles, welcomed at the Osaka Summit in June 2019, were drawn from it. The Compendium lists the 38 OECD members, 12 non-members (most recently Cambodia and Croatia, on 15 May 2026) and the European Union as Adherents; the OECD.AI overview page gives 47 adherents.

Sources

Open the official source https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449

Secondary source: https://oecd.ai/en/ai-principles

Local copy

Verification

History

1. 2026-10-18related: [{"slug": "hiroshima-process-guiding-principles-advanced-ai", "id": "G7-01"}, {" → [{"slug": "hiroshima-process-guiding-principles-advanced-ai", "id": "G7-01"}, {" — two-way link added: AR-01, EG-01, G20-01, G20-02, INT-20, NZ-01, UNESCO-01, ZM-01 link(s) to OECD-012026-10-18-fix-related-links · Claude Code (claude-opus-5-5)
2. 2026-10-18Re-verified against its sources, no change — the same wording quoted again from archive/OECD-01/2026-10-17/official-print.pdf2026-10-18-compare-definitions-4 · Claude Code (claude-opus-5-5)
3. 2026-10-17verification.level: B → A — level computed from the evidence by tools/pipeline.py2026-10-17-recheck-international-1 · Claude Code (claude-opus-5-5)
4. 2026-10-17notes: Adopted 22 May 2019. On 8 Nov 2023 the definition of 'AI system' was revised. A → Adopted on 22 May 2019 (C(2019)34). On 8 November 2023 the Council revised the d — adherents now from the Compendium (38 members, 12 non-members including Cambodia and Croatia on 15 May 2026, and the EU), with OECD.AI's lower count of 47 kept as a discrepancy; the 2024 revision described as the Compendium's background does; the unsupported sentence on the political weight of OECD Recommendations removed2026-10-17-recheck-international-1 · Claude Code (claude-opus-5-5)
5. 2026-10-17summary: The first intergovernmental standard on AI: five values-based principles and fiv → The first intergovernmental standard on AI, adopted by the OECD Council at Minis — Principle 1.4 says AI systems 'should' be safe (a recommendation, not a requirement); the oversight, traceability and risk-management provisions of principles 1.2 b) and 1.5 b)-c), which the topics rest on, added2026-10-17-recheck-international-1 · Claude Code (claude-opus-5-5)

Related records

G7 Hiroshima Guiding Principles G7 · Published (non-binding, current)

HAIP Reporting Framework OECD / G7 · Published (non-binding, current)

Argentina Trustworthy AI Recommendations (2023) Argentina · Published (non-binding, current)

Egyptian Charter for Responsible AI Egypt · Published (non-binding, current)

G20 Johannesburg Leaders' Declaration 2025 (AI section) G20 · Published (non-binding, current)

G20 Innovation Ministerial Statement 2026 (Chapel Hill) G20 · Published (non-binding, current)

HAIP Friends Group Action Plan 2026 Hiroshima AI Process Friends Group · Published (non-binding, current)

NZ Public Service AI Framework New Zealand · Published (non-binding, current)

UNESCO AI Ethics Recommendation UNESCO · Published (non-binding, current)

Zambia National AI Strategy Zambia · Published (non-binding, current)

Citation

OECD Council (at Ministerial level). Recommendation of the Council on Artificial Intelligence [OECD/LEGAL/0449]. 2019-05-22. https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449. Status: Published (non-binding, current); verified 2026-10-17 (level A). Local copy archive/OECD-01/2026-10-11/official.html retrieved 2026-10-11 13:57:39 UTC, SHA-256 2d7447c98e31677dab7c7e4d619f5222ec47160cf1ea91b6b20245efb9b4d3e3. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record OECD-01.

Improve this record

Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.