OECD AI Principles

Claims that cite this source · 9

1. Synthesis· confidence high· passage checked

   Accountability depends on an actor's role, the context of use, and its practical ability to act; it is not a single obligation applied identically to everyone.

   Locator: OECD principle 1.5; NIST AI RMF GOVERN function

   Used in: AI developer or provider, Deployment pressure, Accountability

2. Interpretation· confidence medium· passage checked

   Accountability overlaps with answerability, responsibility, liability and remedy, but those concepts are not interchangeable.

   Used in: Accountability, Contestable decision

3. Synthesis· confidence high· passage review pending

   AI risk management is an ongoing organizational practice across the system lifecycle, not a one-time approval gate.

   Locator: NIST AI RMF core; OECD accountability principle

   Used in: Risk management process, Impact assessment, Pre-deployment evaluation and red teaming, Incident reporting and post-deployment monitoring, Safety

4. Synthesis· confidence high· passage review pending

   Traceability connects decisions and system behavior to records that can support inquiry, review, and accountability.

   Locator: OECD principle 1.5; EU AI Act high-risk obligations

   Used in: Traceability, Audit record

5. Interpretation· confidence medium· passage review pending

   Access to data, compute, expertise, and evaluation capacity shapes which actors can develop systems and which institutions can govern them effectively.

   Used in: Evaluation capacity

6. Interpretation· confidence medium· passage checked

   Adaptability is common: the OECD, the UN advisory body, Japan, Kenya, the African Union and the World Bank all expect governance to be agile or regularly revised. Acting before arrangements set is rarer. Dafoe urges embedding principles while the stakes are relatively low, and Ojanen et al. draw on Guston’s anticipatory governance (“while effective intervention is still possible”), but none of the definitions compared here builds that timing into the definition of AI governance itself. This is the clause in which AIGovSpec departs most from them.

7. Synthesis· confidence high· passage checked

   Of the 29 governmental, intergovernmental and standards documents compared here, only two, Japan’s AI Guidelines for Business and Kenya’s AI strategy, define AI governance in a sentence. Most use the term without defining it, and some laws, including the Council of Europe Convention and Korea’s framework act, do not use it at all.

8. Interpretation· confidence medium· passage checked

   Read in order, the documents move from field-defining scholarship and principles (2017–2021), to organisational management systems and risk functions (2020–2023), to national strategies, laws and standing international bodies (2024–2026). The term spreads faster than any shared definition of it.

9. Interpretation· confidence high· passage checked

   Most recent documents expect governance to adapt, whether described as agile, iterative, living or continually improved, including the OECD, NIST, ISO/IEC 42001, Japan, Kenya, Colorado and Singapore. Far fewer say when governance should act relative to the technology’s spread.