UNESCO Recommendation on the Ethics of AI

UNESCO's General Conference adopted the Recommendation in November 2021. It was negotiated among all UNESCO Member States, which gives it unusually broad legitimacy, especially in the Global South. It rests on four core values: respect for human rights and human dignity; living in peaceful, just and interconnected societies; ensuring diversity and inclusiveness; and environment and ecosystem flourishing. Ten principles follow: proportionality and do no harm, safety and security, privacy and data protection, multi-stakeholder and adaptive governance, responsibility and accountability, transparency and explainability, human oversight and determination, sustainability, awareness and literacy, and fairness and non-discrimination. What sets it apart is its eleven policy areas, which translate values into action: ethical impact assessment; ethical governance and stewardship; data policy; development and international cooperation; environment and ecosystems; gender; culture; education and research; communication and information; economy and labour; and health and social well-being. The text also states that AI systems should not be used for social scoring or mass surveillance. To support implementation, UNESCO developed a Readiness Assessment Methodology that examines a country's AI landscape across legal and regulatory, social and cultural, economic, scientific and educational, and technological and infrastructural dimensions. Dozens of countries have undertaken these assessments. UNESCO also offers an Ethical Impact Assessment tool for evaluating specific systems, particularly in public procurement. A common misconception is that the Recommendation is an enforceable treaty. It is a recommendation: states are expected to apply it and report on progress, but there are no sanctions. Its differences from binding law such as the EU AI Act are scope and force: it covers ethics across society, while binding laws impose specific, enforceable obligations on providers and deployers.