Only 2% have an AI strategy

Only 2% of secondary schools in England have a comprehensive AI strategy, despite AI already being embedded in teaching and learning.

In this episode:

• A new report reveals only 2% of secondary schools in England have a comprehensive AI strategy, despite widespread informal adoption for tasks like lesson planning.
• The primary barrier to implementing AI strategy schools England is a lack of staff confidence and skills (63%), not financial cost.
• School leaders who actively engage with and model AI use foster more consistent AI adoption across their institutions and are key to effective AI staff training schools.
• A clear AI policy education should focus on 'purpose over technology,' defining *why* AI is used, not just *how not to*.
• Uneven AI integration and AI staff training schools, particularly between London and other regions, risk deepening educational disparities in England.

Chapters:

• 00:00 — Cold open & welcome
• 00:25 — Only 2% of schools in England have an AI strategy
• 00:55 — Distinguishing AI policy from AI strategy in education
• 01:30 — Current AI uses for lesson planning and efficiency
• 02:00 — Leadership engagement with AI and perceived risks
• 02:40 — Biggest barriers: staff confidence and AI literacy
• 03:15 — Impact of leadership modelling on AI adoption
• 03:55 — Regional disparities in AI use and widening inequality risks
• 04:50 — Five practical steps for an AI strategy schools England
• 06:50 — Seizing opportunity through purposeful AI strategy

How many schools in England have a comprehensive AI strategy?
Only 2% of secondary schools in England have developed a comprehensive AI strategy, according to a report by Accenture and Teach First, despite AI already being embedded in teaching and learning.

What are the biggest challenges for schools implementing AI?
The primary challenge for schools implementing AI is a lack of staff confidence or skills (63%), followed by data privacy concerns (51%) and a limited understanding of AI's educational potential.

How can school leaders encourage AI adoption among staff?
School leaders can encourage AI adoption by directly engaging with AI, demonstrating responsible use, allowing controlled experimentation, and fostering ongoing shared learning among staff.

Featuring: Dan Fitzpatrick, Accenture, Teach First, Matt Prebble, James Toop, England, London.

Follow AI in Education with Dan Fitzpatrick for more on AI in education.

Transcript

SPEAKER_00 0:00

If this episode makes you think, please let us know in the comments and support us by subscribing and leaving a review. Thank you. Today we are exploring a really eye-opening report from Accenture and Teach First that looks at AI adoption in secondary schools across England, and what it found was pretty stark. Only two percent of the schools they surveyed have actually developed a full comprehensive strategy for how AI should support teaching, learning and school operations. Two percent I found this fascinating because the report makes it clear that AI isn't some far off future thing anymore. According to Matt Preble, who heads up Accenture in the UK and Ireland, AI is already embedded in day-to-day teaching and learning. The challenge he says is that adoption is moving a lot faster than the strategy, training and support needed to sustain it. And I think that's the nub of it, isn't it? Many school leaders are really just trying to navigate this shift without clear guidance, without the confidence to implement the technology effectively. The first thing that jumps out to me from this research is that distinction between an AI policy and an AI strategy. The report tells us that 12% of surveyed schools had an AI policy in place. That's about setting rules, defining what not to do, maybe some safeguard and guidelines, but only 2% had that fully developed strategy. That's a huge difference and it immediately makes me think about one of my core pillars, purpose over technology. It's that idea of starting with why, not how. A policy often focuses on the how not to mess up with the technology, but a strategy, a real one, asks why are we even doing this? What educational purpose does this serve? What do we want AI to achieve? The report highlights that teachers are already using AI for incredibly useful things, like lesson planning, creating quizzes, and drafting exam questions. We're talking about those box one, linear innovations here, using AI to do what we already do, but better. Automating tasks, making things more efficient. And that's fantastic because it aligns perfectly with my philosophy of outsource the doing, not the thinking. The machine handles the repetitive, the first draft, the heavy lifting, freeing up the educator for the real judgment, the creativity, the relationship building. But the piece makes the case that this use is often developing informally, without a consistent, school-wide approach. That lack of a shared model or system for exchanging evidence means activity can be fragmented, and this is where a cohesive AI strategy for schools in England really comes into play. You see, without that clear strategic direction, you're not fully leveraging the potential, nor are you building that collective capability. The research also gives us some numbers on how much leaders themselves are engaging with AI. Sixty percent of school leaders surveyed said they used AI at least weekly, with sixteen percent using it daily. Fourteen percent said they didn't use it at all, and overwhelmingly most leaders were positive about its potential, with 79% agreeing that AI could improve teaching and learning outcomes. That's great, but the interview suggested that many were clearer about the possible risks, like plagiarism, safeguarding, data privacy or reliance on AI generated work, than they were about the specific educational uses. It's natural to focus on what could go wrong, but we also need to build that imaginative muscle for what could go right. This brings me to the second key finding, the biggest barrier to adoption. It's not cost, which was only cited by 16%. The most frequently reported obstacle identified by 63% of leaders was a lack of staff confidence or skills. 51% cited data privacy concerns, and 41% pointed to a limited understanding of how AI could truly support teaching. This really resonates with me. Teachers often get labeled as resistant to change, but more often than not they just need time and space, and they become the best drivers of innovation. We can't just throw new tools at educators and expect them to figure it out. We need to build their no-like trust in these new technologies and in their own ability to wield them. This isn't about teaching coding, it's about building AI literacy as a collaborative reasoning ability, understanding AI limitations, managing conversations with precision, exercising theory of mind. It's about teaching students not to outsmart machines, but to outthink them, and educators need to lead the way in that. Now here's where it gets interesting regarding leadership. The report found that AI adoption was far more consistent in schools where senior leaders actually used AI themselves and demonstrated how it could be applied. Unsurprisingly, schools with more skeptical or disengaged leadership had slower, less coordinated adoption. This isn't about mandating use, it's about modelling it, it's about building champions. If leaders are engaging, if they're experimenting, if they're sharing what they've learned, that enthusiasm starts to spread organically. It's a fundamental principle of change leadership. We also see a regional difference here. In London, 29% of surveyed school leaders said they used AI every day, compared with just 12% across the rest of England. And this is a really important point. While the study doesn't establish whether London schools use AI more effectively or if students receive greater benefit, Accenture and Teach First warn that this uneven access to staff development and technology could widen existing differences between schools and regions. James Toop, Chief Executive Officer of Teach First, rightly points out that there's a real risk that variable access to technology and training could deepen existing inequalities between schools and pupils, particularly outside of London. We need to remember that AI should be an equalizer, not something that creates more disparity. We should be using AI to bridge gaps, especially for the often invisible middle 80% of students, those who aren't at the top or needing intensive intervention, but could really benefit from personalized support or enhanced differentiation. If you're finding these insights helpful and want to stay ahead in the world of AI in education, remember to follow and subscribe to the podcast for more discussions like this. The report also offers some really practical advice. Five steps for schools to start building a proper AI strategy. First, leaders need to engage directly with AI and demonstrate responsible use. You can't ask teachers to do what you won't do yourself. This isn't about being an expert overnight, but showing a willingness to learn and experiment. Secondly, schools are advised to define their purpose and boundaries before widening access. This is that start with why not how principle again. What activities is AI appropriate for? Where must human judgment remain central? This aligns with my core philosophy of human in the loop. AI outputs are drafts requiring critical review, always. Machines can compute. They cannot wonder, they cannot care. These are human domains we must preserve. The third recommendation is to begin with lower risk uses where the value is easier to identify, while building towards a broader plan. Think about those administrative tasks generating lesson starter ideas, rather than full blown assessment design right out of the gate. This ties into the audit step of my seven lessons framework for AI adoption. Evaluate tools through hands-on pilots, not marketing claims. Fourthly, the report calls for schools to allow controlled experimentation within clear professional and safeguarding limits. Give teachers a sandbox, a safe space to try things out without fear of breaking anything or making a mistake. This is about building dedicated structures like AI Maverick's working groups, as I talk about in the accelerate phase of my framework. And finally, the fifth recommendation is to develop capability through ongoing shared learning, rather than relying only on standalone training sessions. This means staff comparing tools, discussing their mistakes, because mistakes are how we learn, and sharing examples of effective practice. This is the very essence of peer-to-peer learning networks, those coffee cart conversations that drive ground level insights and truly activate teachers as change agents. It helps us avoid creating cognitive debt from over reliance by building a deeper understanding. Ultimately, what this report from Accenture and Teach First tells us is that the genie's out of the bottle. AI is here. Teachers are using it, and school leaders recognize its potential, especially in secondary schools. But the current situation, with only 2% having a full AI strategy, is fragmented and risks widening existing inequalities. The biggest hurdle isn't the technology itself, or even the cost, but building staff confidence and embedding a clear vision and purposeful approach. When we embrace this, we can give teachers back time, focus, and energy so they can deepen their connection with students, which is after all why we're all in this profession. AI strategy for schools in England isn't just about managing risk, it's about seizing opportunity, about empowering educators, and about enhancing the truly human aspects of learning. That's all for today. Thanks for listening.