UNESCO publishes first global guidance on generative AI in education

UNESCO published what it described as the first global guidance on generative AI in education and research, addressed to its 193 member states as ChatGPT and similar tools spread through classrooms faster than any policy framework had been built to govern them.

The document set out seven steps for governments, the most concrete of which recommended a minimum age of 13 for students to use generative AI chatbots without teacher supervision — matching the age threshold ChatGPT itself already applied under US children’s privacy law, though below the 16-year threshold set by the EU’s GDPR for independent consent to data processing. The guidance also called for national or regional data-protection standards specific to generative AI, human-rights-based regulatory frameworks, and structured teacher training, on the grounds that fewer than 10% of schools and universities at the time had any institutional policy governing these tools.

UNESCO framed the guidance as a response to a regulatory vacuum: generative AI tools had been adopted directly by students and staff before any government, and in most cases any institution, had assessed their implications for academic integrity, data privacy or the accuracy of the information they produced. The organisation argued that adoption should follow, not precede, the human-agency and child-protection safeguards it set out — the inverse of how the technology had actually spread.

As a UN specialised agency, UNESCO’s guidance carried no binding force; implementation depended entirely on national governments choosing to act on it. It nonetheless became a frequently cited reference point in the national and institutional education-AI policies that followed over the next two years, functioning as an early baseline against which individual countries’ rules on chatbot use in schools were measured.