The OECD AI Principles, adopted in May 2019, were the first intergovernmental standard on artificial intelligence. They were agreed by the member countries of the Organisation for Economic Co-operation and Development and a group of non-member adherents, and later endorsed in substance by the G20. The OECD describes the principles as a guide for building AI that is “trustworthy” and that “respect human rights and democratic values.” Because they predate the EU AI Act and most national AI laws, the principles became a common reference point that later frameworks borrowed from.
The standard sets out five values-based principles: inclusive growth, sustainable development and well-being; human rights and democratic values, including fairness and privacy; transparency and explainability; robustness, security and safety; and accountability. It pairs these with five recommendations to governments, covering investment in AI research and development, building an enabling ecosystem, shaping governance, preparing workers for labour-market transitions, and international cooperation. The OECD updated the text in May 2024 to keep it current with generative AI and other developments.
By the time of the update, 47 countries and entities had committed to the principles, including all OECD members, the European Union, and non-member countries such as Argentina, Brazil, and Singapore. The OECD also runs the OECD.AI Policy Observatory, an online platform that tracks national AI policies, publishes metrics on AI trends, and hosts an AI Incidents and Hazards Monitor, giving governments a shared place to compare approaches built on the principles.