Trends already under way

AI literacy is becoming a curriculum topic in its own right. Pupils are learning how these systems are trained, why they make mistakes and how to use them responsibly. UNESCO has published guidance on generative AI in education (2023) and AI competency frameworks for students and teachers (2024), which schools can use as a reference.

Tools are also becoming multimodal, able to listen, speak, read images and produce diagrams. This could help pupils who find reading difficult or who are learning in a second language, provided accuracy and privacy are checked.

Assessment is another area likely to change. As AI tools make it easy to produce polished written work, many schools are placing more weight on spoken explanation, drafts completed in class and practical tasks. This is less a new invention than a return to methods teachers already know, used more deliberately.

Teaching how AI works, not only how to use it

Pupils understand AI better when they build something simple themselves. Free tools such as Google's Teachable Machine let a class train a small image model in a browser, and Scratch has extensions that link projects to machine learning activities. Combined with physical computing kits such as the BBC micro:bit, pupils see that a model only knows what it was shown.

A classroom activity on training data

Suited to Year 6 (ages 10 to 11) in a 45-minute lesson. Materials: one laptop with a webcam per group, Teachable Machine in a browser, and two sets of objects such as pencils and rulers.

• Groups train a model to tell pencils from rulers using 20 photos of each (15 minutes).

• They test it with new objects, including ones that are neither (10 minutes).

• They retrain using only red pencils and see what changes (10 minutes).

• The class discusses what this shows about biased or limited data (10 minutes).

Preparing as a school

Leaders do not need to predict the future to plan for it. Regular staff training, a clear acceptable-use policy that is reviewed each year, and a small group of teachers who trial new tools and report back give a school a steady way to respond as tools change. Parents value being told plainly which tools are used and why.

Key points

• Treat AI literacy as content to teach, not only a tool to use.

• Use hands-on activities so pupils see how training data shapes results.

• Review AI policies yearly and trial new tools with a small team.

• Keep parents informed about which tools are used and why.

Photos: Deniz Demirci, Jessica Lewis / Unsplash