Ethics

The principles governments, institutions and companies set for AI.

More than 60 countries, China among them, signed the Paris AI Action Summit declaration in February 2025; the US and UK declined. More than 70 countries are running UNESCO's AI ethics readiness assessment.

Countries backing AI ethics frameworks: UNESCO Recommendation adopters 193; UNESCO readiness assessments 70+; Paris summit declaration 60+; OECD AI Principles adherents 47 COUNTRIES BACKING AI ETHICS FRAMEWORKS UNESCO Recommendation adopters 193 UNESCO readiness assessments 70+ Paris summit declaration 60+ OECD AI Principles adherents 47 COUNTRIES
UNESCO readiness countries
70+
Paris declaration signatories
60+
Chinese bodies issuing ethics review rules
10
Major powers not signing at Paris
2

Evidence

Ethics frameworks multiplied while the US and UK stepped back from joint declarations

Documented cases: ethics
DateCaseCountryWhat happenedSource
May 2024OECD AI Principles revisedOECDUpdated for generative and general-purpose AISource
Jan 28, 2025Antiqua et NovaVaticanDoctrinal note on AI and human intelligenceSource
Feb 2025Paris AI Action Summit declarationFranceChina signed; the US and UK did notSource
Jun 2025UNESCO Global Forum, BangkokThailandMore than 70 countries in the readiness programSource
Oct 8, 2023Trial Measures for Ethics Review of Science and TechnologyChinaEthics committees required for sensitive research, including autonomous decision systemsSource

Rules

Ethics frameworks are voluntary except China's ethics review measures

Rules and frameworks: AI ethics
RuleJurisdictionWhat it requiresStatus
UNESCO Recommendation on the Ethics of AIGlobalReadiness and ethical impact assessmentsAdopted Nov 2021
OECD AI PrinciplesOECDTrustworthy AI principlesRevised May 2024
Trial Measures for Ethics ReviewChinaEthics committees and expert re-reviewIssued Oct 8, 2023
Paris Statement on Inclusive and Sustainable AIGlobalVoluntary commitmentsFeb 2025

Status colors: green in force, blue voluntary standard or framework, amber adopted and phasing in, gray draft or pending, red withdrawn.

Record

Page Record

Datasets
ai_trust_evidence.csv (5 rows); registry incidents (0 linked)
Key sources
OECD AI Principles revised
Antiqua et Nova
Paris AI Action Summit declaration
UNESCO Global Forum, Bangkok
Trial Measures for Ethics Review of Science and Technology
Last verified
October 2026