We ran two working parties — the first exploring what AI could do pedagogically, the second building the policies to govern it.
We brought in educational research alongside our technology team, because the question we kept asking wasn't 'what can AI do?' — it was 'what does our community need?'
That's the difference between adopting technology and curating it. And that distinction matters enormously.
Our younger students access AI in a fully monitored, school-controlled environment. They cannot stray.
Our seniors use Gemini through our enterprise Google Workspace — the same grade of data protection as a professional organisation.
In both cases, your child's data never leaves systems we control. We are not using consumer apps. We are not using free tools with unclear terms.
Every piece of work a student receives will be assigned one of these five levels by their teacher.
At one end, no AI whatsoever.
At the other, AI as a creative collaborator — but the student retains full intellectual responsibility throughout.
Notice that even at the highest level, the thinking belongs to your child. AI extends it — it does not replace it. Parents sometimes ask whether AI means students stop working. This is our answer to that question. The work doesn't disappear — it changes shape.
Our biggest risk with AI in schools isn't data privacy — it's what happens to learning when a student can skip the struggle.
The struggle is the learning.
So we've designed our AI use around the Socratic method: the AI asks questions, the student reasons. It will not write your child's essay. It will ask your child what they think the argument should be.
AI removes barriers to understanding — it must never remove the effort of understanding.
In Computing, the hard work is done and we are already running it. Students right now are studying how AI systems actually work — not just how to use them, but how they are built, where they go wrong, and how to evaluate what they produce critically. That curriculum has been substantially redesigned and it is live this year.
Beyond Computing, we are in active curriculum mapping across departments — and the plan is clear. From September, purposeful AI integration will be embedded across the school, with every task mapped to the Scale above.
Our AI framework has been independently reviewed by 9ine — specialists in technology governance for international schools.
Every tool in use has passed their assessment and our school’s DPO before a student has access to it.
Data protection, risk assessment, policy — all of it reviewed against international standards.
For our younger students, we use a 'sandbox'. It’s safe, it’s monitored, and it’s age-appropriate.
This is a Gemini Gem — a custom AI tutor we built inside our Google Workspace that is shared with our students.
It has one rule: never give the answer.
A student asks about river velocity for their GCSE Geography. Rather than explaining it, the AI asks the student to reason about friction first.
That student will remember this because they worked it out.
That's what distinguishes purposeful AI from passive AI.