Should the government run an AI Tutoring Tools Pioneer Programme to co-design tools for disadvantaged pupils?

Britain-on-the-platform could barely separate itself on whether Whitehall should co-design AI tutoring tools with private companies, No prevailing by a single vote out of 45 cast, but the closeness of the top line masks a sharper divide underneath. This was not consensus withheld by indifference; it was consensus withheld by disagreement that cuts along both gender and age lines.

Men rejected the scheme by close to two to one, while women leaned the other way, splitting roughly three to two in favour, a gender gulf as wide as any this category has produced. Age tells its own story: the youngest cohort, presumably closest to the classroom experience of AI tools, backed the programme almost unanimously, but the 35-44s, the demographic most likely to have children in the system now, turned against it hardest of all, with only one in eight voting yes.

The result sits squarely on the fault line of tech-in-public-services scepticism — the same unease that dogs NHS data contracts and exam algorithms, where the promise of scale for disadvantaged pupils collides with distrust of handing curriculum design to eight hand-picked private firms. That the most sceptical group was the one raising school-age children suggests this is less an anti-AI reflex than a parents' verdict on who gets to shape what their children are taught.

Where the DfE hoped for optimism about reaching 450,000 pupils, it got instead a coin toss decided by the people closest to the classroom door.