AI literacy for multi-academy trusts: how to deliver it consistently across all your schools

The MAT AI literacy problem - and why it persists

Research published by the Education Policy Institute in January 2026 made a finding that will be familiar to anyone working in trust leadership: some multi-academy trusts have mandated that all their schools provide AI literacy training for pupils, but it is not happening in practice.

This is not a failure of intent. Trust leaders understand that AI is already shaping the world their pupils are heading into. They understand that the national computing curriculum - unchanged since 2014 - has nothing to say about it. They understand the urgency. The problem is not awareness. The problem is delivery.

AI literacy for pupils is one of those areas where the gap between a trust-level decision and consistent school-level reality is particularly wide. And the reason for that gap is straightforward: it depends almost entirely on individual teacher confidence - which, right now, is not there.

Why building teacher AI confidence alone won't solve it

The instinctive response from trusts is to train their way out of the problem. CPD sessions, DfE modules, access to resources. All of these help. None of them solve the delivery problem at pace.

Teacher AI confidence is a legitimate long-term investment. But it takes time - time to train, time to embed, time to translate into classroom practice that pupils actually experience. A trust that wants consistent AI literacy delivery across all its schools in September 2026 cannot get there through staff training alone. There is not enough runway.

There is also a harder truth here. Even where teachers have engaged with AI tools personally, teaching pupils about AI - what it is, how it works, what the risks are, what the law says - requires a different kind of knowledge and confidence than using Copilot to write a lesson plan. The Royal Society has noted that most existing AI education efforts focus heavily on technical skills with less emphasis on the broader societal and ethical dimensions. That is the content a Year 6 pupil or a Year 9 class actually needs.

The AI literacy gap between state schools and independent schools

The argument for trusts to act is not just about compliance or curriculum. It is about what happens to pupils who do not get this.

Research from the Tony Blair Institute found that independent school teachers are 2.5 times more likely to teach pupils how to use AI in their learning, and 3.5 times more likely to teach how AI applies within subjects. Independent schools are three times more likely to have a school-wide AI strategy. The Sutton Trust's Artificial Advantage report corroborates this, finding that the AI literacy gap between independent and state schools is already opening a new dimension of educational inequality - one that multi-academy trusts are uniquely positioned to close across their schools.

The pupils in state schools - including the schools in your trust - are falling behind on AI literacy before they have even started. The gap is already opening. And it compounds: the children who understand AI at 11 will be better equipped to engage with it at 16, and at 21, and in the jobs market where mentions of AI literacy in professional job descriptions have risen 69% since late 2024, according to Tony Blair Institute analysis of LinkedIn data.

For a trust committed to raising outcomes and closing gaps, the AI literacy gap is a serious equity issue. It is also one that trust leadership can actually do something about - now, not in 2028 when the reformed curriculum arrives.

What consistent AI literacy delivery across a MAT actually requires

The trusts making progress on AI literacy share a common approach. They are not waiting for teacher confidence to catch up. They are separating the question of what pupils need from the question of who delivers it.

At the trust level, that means being clear about what AI literacy looks like at each key stage. Not a vague aspiration - a defined outcome. By the end of KS2, every pupil should be able to answer: what is AI, where does it already exist in my life, and what are the risks I need to be aware of? That is a testable, deliverable foundation.

At the school level, it means not placing the burden of delivery entirely on class teachers who are already stretched and who were never trained for this. The model that works for other sensitive or specialist content - drugs education, sexual health, safeguarding - works here too. External specialists come in, deliver a credible session, and leave teachers better informed for the conversations that follow.

A practical AI literacy framework for MAT and trust leaders

If you are responsible for AI strategy across a trust and you want consistent pupil-facing AI literacy delivery by September, here is what that looks like in practice.

Define the outcome, not the method. Start with what you want every pupil to know at each key stage. Keep it simple. A Year 6 pupil who leaves primary school understanding what AI is, where it exists, and that it can be used to create content that is not real - that is a foundation. Build from there.

Separate staff AI use from pupil AI literacy. These are related but different. Teacher CPD on using AI tools is valuable. It is not the same as delivering age-appropriate AI literacy to pupils. Both matter. Do not conflate them in your planning or your budget.

Use external specialists for consistency and pace. A trust-commissioned programme delivered by an external provider gives you consistent content across all your schools without depending on individual teacher confidence. It scales in a way that internal delivery currently cannot. It also gives you something to point to - a named programme, a delivery record, evidence of provision.

Start with Year 6. The practical entry point for most trusts is the Year 6 cohort. They are about to enter secondary school, where AI literacy and the RSHE deepfakes requirement both become more pressing. A 20-minute assembly that gives an entire year group a shared foundation is a high-impact, low-disruption starting point. It requires nothing from staff beyond booking it.

Map your secondary compliance gaps now. If your trust includes secondary schools, check which ones have RSHE provision in place for the September 2026 deepfakes requirement. The ones that do not are your most urgent need. External delivery for those sessions can be confirmed and dated before term ends.

The consistency problem is real. But it is not unsolvable. The trusts that will have genuinely consistent AI literacy across their schools by the end of 2026 are the ones that stopped waiting for the right conditions and started commissioning delivery. The conditions are not going to get easier. The urgency is only going to grow.

For more on the curriculum gap that makes this urgent, read our piece on what the national curriculum currently says about AI - and what that means for the pupils moving through your schools right now.