The scale of the bot flood

When Rishi Sunak's Conservative government consulted in 2023 on plans to ban disposable vapes and create a smoke-free generation, more than three-quarters of the submissions it received were not from people at all. The Department of Health and Social Care's bot-detection methods identified 90,835 of the 118,756 responses - 76% - as machine-generated, in what researchers described as an attempt to derail the disposable vapes ban uk.

The automated campaign had a recognisable pattern, according to the DHSC:

• All but two of the 90,835 automated responses answered the same question, opposing the plan to ban disposable vapes.
• The submissions were often duplicated many times.
• They were sent at high frequency from a small number of IP addresses.

How it was found

The manipulation was disclosed in research published in the journal Tobacco Control by Dr Nason Maani, of the University of Edinburgh, and Prof Anna Gilmore, of the University of Bath. The researchers said the bombardment appears to be the first documented instance of large-scale automated interference in a formal public health policy consultation.

What happens now

The flood failed to change the outcome. The DHSC excluded the 90,835 automated responses from its decision-making, and the bill - inherited by Keir Starmer's Labour government in 2024 - passed into law with the ban on selling or supplying disposable vapes included.

A DHSC spokesperson said the department's defences held. "We always have robust controls in place to prevent the manipulation of our public consultations, and those controls worked effectively in this case." The spokesperson added that fake responses were removed and that the department will remain alert to future interference.

Warnings for future consultations

Dr Maani warned that the technique could get harder to detect over time. "the threat is that future campaigns become harder to catch, and what we can't catch, we don't know about", he said, arguing that such campaigns could make a handful of powerful interests look like a mass movement.

Campaigners and experts on democratic process said the episode exposed a wider vulnerability. Caroline Cerny, deputy chief executive of Action on Smoking and Health, said: "Evidence that AI bots have been used to swamp a Department of Health consultation is deeply concerning." She called for safeguards, saying decisions that affect public health must rest on sound evidence and expertise rather than manipulated response numbers.

Azzurra Moores, associate director for information ecosystems at the thinktank Demos, said the findings are pretty alarming, because a consultation that can be swamped stops telling policymakers what the public actually thinks. She added that the problem has no single fix: every barrier that keeps bots out also makes participation harder for real people, and AI is making fake responses harder to spot.