AI Why you need AI to effectively deliver the SEND reforms
TIM HANDLEY, former teacher, data lead and primary headteacher and now a certified Google AI practitioner and developer of Edu Intelligence’s sector-leading 22 specialist agent education- specific AI, shares his insights.
The SEND reforms in the Schools White Paper are the most ambitious overhaul of provision since 2014: a three-layer support model; a reformed, more strategic SENCO; Digital Individual Support Plans for every child with identified SEND; and nationally defined Specialist Provision Packages.
As a former school leader, I want to see this delivered, but the potential workload burden is high, meaning trusts and schools need to embrace safe and secure AI to help. Every child with identified SEND will need a digital ISP, reviewed annually, and inspected as part of Ofsted visits. Every school must publish an Inclusion Strategy backed with evidence of the impact delivered, and accounting for how it deploys its inclusion funding. Ofsted is already grading Inclusion as a standalone area, with the Strong standard explicitly requiring “well-analysed, quantitative and qualitative data” to underpin leaders’ decisions. Together, the reforms create the biggest increase in SEND data, tracking and evidencing requirements the sector has seen in a generation. The honest conversations happening in SENCO offices and trust SEND leads’ meetings right now is that leaders cannot deliver what the reforms ask using the systems and methods they have today. Not for lack of effort or commitment, but because of the workload increase.
You cannot hand a SENCO a more strategic, less administrative role while mandating digital ISPs for every SEND child. You cannot expect a school to publish an evidence-based Inclusion Strategy when assembling the evidence takes a term of manual cross-referencing across multiple systems and documents.
You cannot hit the Strong standard for inclusion on “well-analysed, quantitative and qualitative data” if this depends on one person’s evenings and a spreadsheet. And you cannot make SEND groupings meaningful if
the only way to share insight across schools is PDFs stitched together from exports.
This is where AI stops being a conference topic and moves to being the delivery mechanism. Not classroom AI, but leadership AI. AI that sits across the SEND data a school already holds, including EHCPs, ISPs, Ed Psych reports, assessments, attendance, and behaviour, and answers the questions leaders need answered: Is this provision being delivered? Is it working? Where is the gap widening? Which children need something different, and why?
The White Paper’s own language; insights “put directly into the hands of leaders” describes exactly this. It cannot mean another dashboard. Dashboards show what happened; they don’t explain why, and they don’t recommend what to do next. SENCOs and inclusion leaders don’t need more information. They need analysis, context, and time back. Two things matter. First, it must be education specific. Generic AI doesn’t understand graduated response, EHCP provisions, or how SEND cohorts compare across year groups. It can sound plausible about schools while being wrong about them, and with vulnerable children, wrong is not good enough. Second, it must be safe. Walled-garden, GDPR-compliant, UK-processed, and designed so the AI can surface connected insight without ever needing to see identifiable student data.
The SEND reforms describe a system that as a sector we want every child to experience. But delivery won’t happen through policy intent alone. It will happen when every SENCO, inclusion leader and headteacher has AI doing the connective, analytical work that currently eats their week, so the strategic leadership the reforms envisage becomes possible.
Without it, the reforms risk becoming another layer of expectation on leaders already running at capacity. With it, we might finally deliver what children with SEND and their families have been promised for a decade.
Striking the balance? Education Today hears from cyber security and tech hub, CyNam.
Artificial intelligence has arrived, not as a far-off concept, but as a present and growing influence on how learning is delivered, assessed and supported. As schools and colleges navigate this shift, a recurring question emerges: is there still time to shape AI’s role with confidence, care and – perhaps – legislation?
The truth is, the answer depends on who you are. Senior leaders balancing budgets, curriculum leads driving innovation, IT managers safeguarding systems and classroom teachers managing mounting workloads will all see AI’s arrival through very different lenses. At a recent roundtable organised by CyNam in partnership with HCR Law, well attended by regional academic decision makers, I was struck by the enormous gap between AI adoption and the emotions it evoked. Recent data from the Sutton Trust (July 2025) shows that over half of teachers have already incorporated AI in some form, from lesson ideas and resource drafting to, in some cases, marking. Yet many use it quietly and hesitantly. This reluctance raises questions about training, leadership direction and professional culture. Adoption has been uneven, largely left to confident individuals, and notably more prevalent in independent schools than state settings.
While educators deliberate, students are already experimenting, sometimes skilfully, sometimes clumsily and occasionally inappropriately. This gap between teacher confidence and student experimentation creates a new challenge: how do we guide learners in ethical, transparent use of tools they are already reaching for? Pandora’s box has been opened. AI is not going away. Our task now is to understand it, shape it and position it
May 2026
as a complement to human expertise, not a replacement for it. AI can draft, suggest and automate, but it cannot empathise, inspire or hold expectations for behaviour, ambition or wellbeing. These remain fundamentally human responsibilities. Think of AI as an apprentice: capable of completing tasks with speed and precision, but still needing boundaries, review and correction.
Used well, AI has real potential to reduce workload, personalise learning, increase accessibility, remove social and linguistic barriers and free up time for the human connections that matter most in the classroom. Used poorly, it can confuse, mislead and erode the very skills we are trying to build Every industrial revolution has disrupted the status quo. Our task now is not to halt this one, as it is already too late for that, but to steer it. That means advocating for clear legislation, ensuring AI is adopted equitably across all school settings and involving governors, trustees and parents in its introduction and governance.
Learning should be challenging, exciting and rewarding. Growth comes from trying, failing and trying again. AI should not do the thinking for learners; it should widen their horizons, deepen understanding and build the confidence to enquire further.
AI’s presence in education is inevitable, but its impact is not yet decided. We have a genuine opportunity to shape how it enriches learning and improves outcomes, particularly for those who previously had little chance to change their circumstances.
The future of learning will not simply be AI-powered. It will be a partnership and it is everyone’s responsibility to make that a success..
www.education-today.co.uk 23
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48
