How Schools Can Use Generative AI Without Losing the Learning

What happened?

The challenge for schools is not deciding whether AI exists. It is designing learning so that students use powerful tools without outsourcing the thinking education is meant to develop. The OECD's Digital Education Outlook 2026 says generative AI can support learning when it is guided by clear teaching principles.

Key points

• The OECD's Digital Education Outlook 2026 says generative AI can support learning when it is guided by clear teaching principles.
• The OECD and European Commission have published an AI literacy framework for primary and secondary education.
• OECD policy work on higher education identifies academic integrity, data protection, reliability and equity as key concerns.
• UNESCO continues to emphasize responsible AI governance and human-centred approaches to technology.
• Schools are increasingly moving from blanket bans toward clearer rules about when AI assistance is appropriate.

What we know

1. The OECD's Digital Education Outlook 2026 says generative AI can support learning when it is guided by clear teaching principles.
2. The OECD and European Commission have published an AI literacy framework for primary and secondary education.
3. OECD policy work on higher education identifies academic integrity, data protection, reliability and equity as key concerns.
4. UNESCO continues to emphasize responsible AI governance and human-centred approaches to technology.
5. Schools are increasingly moving from blanket bans toward clearer rules about when AI assistance is appropriate.

What officials/people involved say

OECD research stresses that educational value depends on how AI is used. UNESCO's broader AI-governance work similarly emphasizes human rights, responsibility and informed use.

Background

Generative AI can explain concepts, provide examples, translate text, simulate dialogue and offer feedback. But it can also produce incorrect information and allow students to submit work they did not meaningfully understand.

Why it matters

The answer is not to pretend AI is not available. Students will encounter it in university, employment and everyday life. Schools therefore need to teach students how to question AI, verify sources, protect personal data and disclose assistance when required.

What happens next?

A practical policy can define three zones: AI prohibited for work intended to measure unaided mastery; AI permitted as a learning assistant with disclosure; and AI encouraged for brainstorming, accessibility or productivity. Assessment should also include oral explanation, process evidence and supervised work where appropriate.