More than 300 teachers, parents, academics and doctors are calling for a pause on plans to use AI tutoring tools in schools, as the Government says it is running a controlled trial, not a national rollout. The dispute matters because the proposal could reach up to 450,000 children from disadvantaged backgrounds, but the reported account gives no evidence yet that the tools improve learning.
Watch Desk analysis
What happened
The Mirror reports that signatories to a letter addressed to Greater Manchester mayor Andy Burnham and Education Secretary Lucy Powell want plans for AI tutoring to be paused. The article says up to 450,000 children from disadvantaged backgrounds could be offered the tools from next year, with the stated aim of improving access to tutoring that many families cannot afford.
The signatories argue that the technology is unproven and that children’s needs, including food poverty and mental health support, cannot be solved by technology. Dr Velislava Hillman, a co-author of the letter, called for the Government to invest in established support and only return to AI tutoring if it can show more benefit than harm.
A Government spokesperson told the Mirror this is not a national rollout. They described it as a carefully controlled trial intended to build evidence about what works and make sure the tools are safe before any wider expansion. The article does not specify the trial’s design, participating schools or how its results will be assessed.
Why it matters
AI tutoring is being presented as a possible way to widen access to extra help. Whether that promise holds depends on what the tools can actually do for pupils, how their effectiveness and safety are tested, and whether the trial measures outcomes that matter to children and teachers.
The distinction between a trial and a national rollout is important: the Government says the immediate plan is to gather evidence, while the letter’s signatories warn against placing the risks on children with the fewest alternatives. Both claims point to the same practical question: what evidence would justify moving beyond a limited test?
Our read
This is a real policy argument, not proof that AI tutors are either a fix for educational inequality or a danger to pupils. The Government’s response describes a trial; the published account does not give enough detail to judge its safeguards or how success will be measured. Those details should be public before trial results are used to make the case for wider adoption. Education is not short of grand promises, and children should not have to serve as the small print.
What to watch
- Which schools and pupils will take part, and what safeguards will apply.
- How the trial will measure learning, safety and effects on teachers’ work.
- Whether the results and evaluation methods will be published before any expansion.
Discussion spark: What evidence should a school AI tutoring trial have to show before it can expand: better learning outcomes, proven safety, or both?
Sources and evidence
- Source update (8 November 2026, 17:16 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.