Should the UK allow generative AI to support licensing and ethical oversight in animal research?

There was no ambiguity here: the closest the No side came to a contest was among the 35-44s, who alone leaned narrowly for AI's involvement, but every other age band, from the youngest voters to pensioners, dug in against it, with the under-25s coming closest to unanimity in their refusal.

The gender gap was just as telling. Men split closer to evenly, roughly two-fifths open to the idea, while women rejected it by more than five to one — a divide that tracks a familiar pattern in these debates over automation touching welfare and safeguarding, where women's scepticism of algorithmic delegation tends to run deeper than men's.

The result sits squarely inside the wider unease about AI creeping into regulatory and ethical decision-making, echoing fights over algorithms in benefits assessments, sentencing tools and clinical triage: voters seem willing to accept AI as an administrative aid but not as a proxy conscience, especially where the subjects of a decision cannot object for themselves.

That instinct found its sharpest phrasing in the top-rated No argument, from a voter posting as "Rabbit," who insisted AI "should have no responsibility over any decision that involves anything more than bit routing" — a line that drew the loudest applause on the floor, with no comparably endorsed case emerging for the other side.

However useful the Home Office believes generative AI might be for paperwork and pattern-spotting, this chamber wants a human hand, not a model, on anything resembling a moral judgement.