OECD AI Principles

Adopted in 2019 and updated since, the OECD AI Principles were the first intergovernmental AI policy framework — and more than 45 countries, including every jurisdiction we track here, have formally adhered to them. They're not directly actionable the way the NIST AI RMF or ISO/IEC 42001 are; they operate one level up, as the shared values baseline that national AI laws and frameworks get built on top of.

The five values-based principles

The Principles commit adhering countries to AI that is: beneficial and works for people and the planet; respects the rule of law, human rights, and democratic values including fairness and privacy; is transparent and subject to appropriate disclosure; is robust, secure, and safe throughout its lifecycle; and holds organizations and individuals accountable for its proper functioning.

Why a non-binding, values-level framework still matters practically

You won't build a compliance checklist directly from the OECD Principles — they're too high-level for that. What they're actually useful for is context: when you see the EU AI Act's risk-tiered structure, Colorado's duty of reasonable care, or the UK's five cross-sectoral principles, you're looking at different jurisdictions' concrete implementations of substantially the same underlying values commitment. Understanding the OECD Principles makes it easier to recognize what's genuinely novel in a new jurisdiction's AI law versus what's a familiar principle wearing different statutory language.

Practical relevance for multinational compliance programs

For a company building one AI governance program that has to satisfy multiple jurisdictions, the OECD Principles are a useful sanity check: a program that genuinely satisfies all five principles in substance is unlikely to be caught flat-footed by a new jurisdiction's AI law, precisely because most new AI legislation is a more specific, binding implementation of principles the OECD framework already articulated at a higher level.