OECD AI Principles

Why it matters

They are the closest thing AI governance has to a shared starting vocabulary - "human-centred values," "transparency," "robustness" - that later, more binding instruments (the EU AI Act, national strategies, UNESCO's Recommendation) all build on rather than contradict.

Key takeaways

• Five values-based principles: inclusive growth, human-centred values and fairness, transparency and explainability, robustness and safety, and accountability.
• Five recommendations for policymakers, covering national AI strategies, investment in R&D, an enabling policy environment, building human capacity, and international cooperation.
• Revised in May 2024 to explicitly address generative AI and foundation models, which did not exist in their current form when the 2019 text was adopted.
• Underpin the G20 AI Principles, adopted the same year, extending the same language to non-OECD G20 members.

Part of these reading paths

• Understanding AI Governance