Teaching AI Ethics to 10-Year-Olds: Lessons from Our First Cohort

Starting with Fairness, Not Functions

When most schools introduce artificial intelligence, they begin with the technical: algorithms, data structures, perhaps some introductory coding. We took a different path entirely. We began with a story about a robot that had to decide who gets the last seat on a lifeboat.

The room erupted. Ten-year-olds, it turns out, have incredibly nuanced views about justice, bias, and the limits of machine decision-making — if you give them the space to explore.

The question isn't whether AI will shape our children's world. It's whether our children will be equipped to shape AI.

A Curriculum Built on Questions

Our AI literacy programme is structured around essential questions rather than technical competencies. Questions like: "Can a computer be creative?", "Who teaches the machine, and what biases do they bring?", and "Should an AI ever be allowed to make decisions about people?"

These questions lead naturally into technical concepts — training data, bias, neural networks — but the entry point is always human. Always ethical. Always grounded in the child's own experience of fairness and power.

What We Learned

Three things surprised us in the first cohort. First, children's capacity for philosophical reasoning about technology far exceeded our expectations. Second, the emotional dimension of AI ethics — fear, wonder, empowerment — was just as important to address as the intellectual dimension. Third, parents became deeply engaged, leading to some of the richest family learning conversations we've seen at the school.

We are now developing a published framework for AI ethics in primary education, and we look forward to sharing it with the wider education community later this year.