Summary
- Adopted by the OECD Council on 22 May 2019. Revised on 8 Nov 2023 (new definition of "AI system") and again on 3 May 2024 (updates for generative AI and easier implementation).
- Five values-based principles for all AI actors (section 1):
- Inclusive growth, sustainable development and well-being
- Respect for the rule of law, human rights and democratic values, including fairness and privacy
- Transparency and explainability
- Robustness, security and safety
- Accountability
- Five recommendations for governments (section 2): invest in AI R&D; foster an inclusive AI-enabling ecosystem; shape an interoperable governance and policy environment; build human capacity and prepare for labour market change; and cooperate internationally on trustworthy AI.
- It is the source of the G20 AI Principles (welcomed at the Osaka Summit, June 2019). Its AI system definition is widely reported to have shaped later laws such as the EU AI Act (secondary observation, not stated in the OECD source).
- Non-binding: the OECD itself explains that Recommendations carry political weight and an expectation of best efforts, not legal force.
Who it applies to
- Governments that adhere (OECD members plus non-member adherents; 47 in total per oecd.ai). It "calls on" all AI actors to apply the principles, but creates no duties for companies.
- Not extraterritorial.
Key dates
- Adopted by the OECD Council at Ministerial level
- G20 Leaders welcome the G20 AI Principles, drawn from the Recommendation (Osaka Summit)
- Council revises the definition of AI system
- Council at Ministerial level revises the Recommendation, including for generative AI
Full date record
| Date | Event |
|---|---|
| 2019-05-22 | Adopted by OECD Council at Ministerial level |
| 2019-06 | G20 AI Principles welcomed at Osaka, drawn from the Recommendation |
| 2023-11-08 | Definition of "AI system" revised |
| 2024-05-03 | Recommendation revised at Ministerial Council Meeting |
Penalties and enforcement
- None. The OECD Digital Policy Committee, through its Working Party on AI Governance, monitors implementation and reports to Council. The 2024 text asks for a report no later than five years after the revision, then at least every ten years.
What it means for an Indian company
- No legal duty. But the five principles are the common vocabulary behind most AI laws and frameworks, including India's own NITI Aayog Responsible AI for All principles and the India AI Governance Guidelines.
- A simple way to show alignment: map your AI policy (SC-AI-001 AI governance policy approved by the board or top management) to the five principles, then point to controls for each, such as SC-AI-024 Explain AI-assisted decisions (transparency), SC-AI-018 Robustness and adversarial testing (red teaming) (robustness), SC-AI-017 Fairness and discrimination testing of outputs (fairness) and SC-AI-002 AI roles, accountability and oversight committee (accountability).
- Useful in sales to OECD-country customers who reference the principles in vendor questionnaires.
Obligations
Full obligation breakdown coming
We break the most requested policies down into numbered obligations mapped to controls. Leave your details and we’ll tell you when OECD AI Principles is done.
Glossary
- OECD: Organisation for Economic Co-operation and Development, a forum of mostly high-income countries that agrees common policy standards.
- Recommendation: an OECD legal instrument that is not binding but signals political commitment.
- Adherent: a country (member or non-member) that has formally signed up to an OECD instrument.
- G20 AI Principles: the version of these principles welcomed by G20 leaders in 2019, which India supports.
- AI actor: anyone who plays a role in the AI lifecycle, including developers and deployers.
Key takeaways
- Five principles plus five government recommendations; non-binding; revised 2023 and 2024.
- India is not an adherent to the OECD Recommendation; it backs the G20 AI Principles derived from it.
- Original adoption was 2019-05-22.
- Best used as the values layer when writing a company AI policy.
Official sources
- OECD Recommendation on AI, OECD/LEGAL/0449 (legalinstruments.oecd.org; print PDF via the site), accessed 2026-10-19
- OECD.AI: AI Principles and list of adherents, accessed 2026-10-19
This is general information, not legal advice. Last verified 19 Oct 2026 against the official source. Laws change; confirm the current position with qualified counsel before acting.