OECD AI Principles

The OECD AI Principles (2019, updated 2024) are “the first intergovernmental standard on AI” — the soft-law convergence layer beneath the regimes this wiki tracks, and the answer to a big chunk of the “whose standards win?” open question: the eu-ai-act, the Council of Europe treaty, US policy, and UN guidance all build on the OECD’s definition of an AI system and its lifecycle. 47 adherents (all OECD members, the EU, plus Brazil, Singapore, …). Source: oecd.ai.

The five values-based principles

Inclusive growth/well-being · human rights & democratic values (fairness, privacy) · transparency & explainability · robustness, security & safety · accountability — plus five recommendations to policymakers (R&D investment, ecosystem, interoperable governance, human-capacity/labour transition, international cooperation).

The definition everyone borrows

The substrate this wiki keeps pointing back to is one sentence. The OECD defines an AI system as “a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments” (oecd.ai). The OECD records that the EU, the Council of Europe, the United States, and the United Nations have incorporated this definition and the OECD AI lifecycle into their own frameworks — which is why the eu-ai-act and the CoE treaty line up at the definitional layer even where their instruments diverge. The May 2024 update revised the principles to keep them “robust and fit for purpose” against newer technology (notably generative AI) and policy developments.

Why it matters here

It’s explicitly the standard a light-touch state like Israel aligns to (responsible-innovation) while only watching the eu-ai-act — “a foundation for global interoperability between jurisdictions.” But it’s non-binding soft law: it guides, it doesn’t enforce. It and iso-iec-42001/nist-ai-rmf form the voluntary-standards tier; the binding tier is the eu-ai-act and the framework-convention-on-ai.

Related

ai-governance · eu-ai-act · framework-convention-on-ai · nist-ai-rmf · iso-iec-42001 · risk-based-regulation · responsible-innovation