We have been here before: AI and the edtech graveyard

Adam Sturdee
May 29
3 min read

“AI readiness is now mainstream.” That is the headline from the latest Bett UK Education Priorities Report, and on the face of it, it is good news. Sixty per cent of educators put AI near the top of their concerns. Policies are being written. Working groups are meeting. Schools are, by their own account, getting ready.

I want to offer a slightly uncomfortable thought. Readiness is not the same as change. And in our hurry to look ready, we risk repeating a mistake this profession has made more than once.

We have been here before

Those of us who have spent a while in schools have seen this film. The interactive whiteboard that became an expensive projector. The virtual learning environment nobody logged into. The class set of tablets that lived in a cupboard. Each arrived with a strategy document, a launch, and a genuine belief that this time the technology would transform teaching. Most changed very little, because we bought the tool and skipped the harder work that makes a tool matter.

AI is different in scale and ambition, but the failure mode is identical. It is entirely possible to adopt AI, write the policy, buy the platform, present at the conference, and leave the actual business of teaching and learning untouched. That is not transformation. It is theatre.

The quiet signals in the data

The interesting part of the Bett data is not the AI headline. It is the quieter signals underneath it. Buyers are starting to ask for measurable impact when they choose technology. In the UK, leaders place roughly twice the weight on staff capability that their international peers do. Concern about secure, compliant implementation is rising, and sharply so among trusts, where it reaches 42 per cent.

Read together, those are the priorities of a sector growing up. They are unglamorous. Evidence, capability, security. None of them will earn you applause at a trade show. All of them are the difference between technology that changes practice and technology that gathers dust.

The questions that actually matter

So here is the challenge I would put to fellow leaders. The question is no longer whether we are using AI. Almost everyone is, or soon will be. The questions that matter are harder, and they are about us, not the technology.

What has it changed? Can we show it, with something more than a vendor's slide? Has it lifted the load on our staff, or quietly added to it? Did we invest in our people's capability to use it well, or assume they would work it out? Is it secure, and could we look a parent in the eye and explain what happens to their child's data?

If we cannot answer those questions, we are not AI ready. We are AI busy.

Be boring on purpose

The discipline this moment calls for is, frankly, a little boring. Demand evidence before you buy, and after. Invest in people before platforms, because capability is the thing that turns a tool into a practice. Protect trust as though it were the asset it is, because once a community decides a technology is being done to them rather than for them, you do not win that back easily.

And resist the shiny. The most dangerous position in school leadership right now is not being behind on AI. It is being seduced by it.

I am an optimist about what AI can do for teachers. Used well, it can give every member of staff access to the kind of frequent, specific and supportive development that was simply impossible to provide at scale before. That is a real prize. But it will go to the leaders who refuse to confuse motion with progress, who ask for proof, and who keep their people at the centre of every decision. Readiness was never the goal. Change is.

Adam Sturdee is a senior leader and co-founder of Starlight, the UK’s teacher-first AI-powered transcript-based coaching platform for educators.

His work sits at the intersection of dialogic practice, instructional leadership and responsible AI strategy for schools and trusts.

He recently presented his research on AI-supported coaching at the BERA TEAN Conference 2026: https://www.bera.ac.uk/conference/bera-tean-conference-2026

We have been here before: AI and the edtech graveyard

We have been here before

The quiet signals in the data

The questions that actually matter

Be boring on purpose

Recent Posts

Comments