Bots wrote 76% of responses to UK vape consultation, study finds
By Anika SharmaAI & Innovations Editor, Peepals Tech
Representative image: councilcle / Pixabay
A Tobacco Control study reported on October 28-29, 2026, found 90,835 of 118,756 responses to the UK's 2023 disposable vapes consultation were bot-generated.
What Happened
About three in four responses to a 2023 UK government consultation on banning disposable vapes were generated by automated "bots", a study published in the journal Tobacco Control has found, according to reports on Wednesday, October 28 and Thursday, October 29, 2026. Researchers from the universities of Edinburgh and Bath found that 90,835 of the 118,756 submissions, about 76%, were machine-generated, and that almost all of them opposed restrictions. The Department of Health and Social Care had flagged the suspicious entries at the time and excluded them; once they were removed, 79.6% of genuine respondents supported restrictions on disposable vapes.
Key Facts
- Study: published in the journal Tobacco Control by researchers from the University of Edinburgh and the University of Bath.
- Consultation: the UK government's 2023 consultation on restricting disposable vapes.
- Submissions: 118,756 in total, of which 90,835, about 76%, were machine-generated.
- Pattern: all but two of the automated responses chose the same multiple-choice option opposing restrictions.
- Detection: officials noticed entries arriving at high frequency from a small cluster of internet addresses, with heavily duplicated wording, and used bot-detection screening.
- Result after cleaning: 79.6% of valid respondents supported restricting disposable vape sales; the UK went ahead with a ban on their sale and supply in England.
- Who was behind it: the researchers said the volume and consistency suggest a targeted campaign, but that it is "not possible" to tell who was responsible.
Why It Matters
Public consultations are one of the main ways ordinary people can influence new laws and rules. Governments publish a proposal, invite comments, and are expected to take the responses into account. If a single actor can flood the process with tens of thousands of automated replies, it can make a minority view look like the majority and push officials towards a decision that most real respondents did not want.
In this case the system worked: officials spotted the pattern, removed the fake responses and based the decision on genuine ones. But the researchers describe it as what appears to be the first documented case of large-scale automated interference in a formal public health consultation, and they warn that the threat will grow. As the lead researcher put it, "the threat is that future campaigns become harder to catch, and what we can't catch, we don't know about." Modern AI tools can write unique, natural-sounding text for each submission, which would defeat the simple duplicate-wording checks that caught this campaign.
The findings matter for any government that consults online, including India, where draft rules are also put out for public comment before they are finalised. Strong identity checks, limits on submissions from the same source and analysis of patterns in timing and wording are among the tools that can protect such processes, though each must be balanced against keeping consultations open and easy to use.
| Measure | Figure |
|---|---|
| Total submissions | 118,756 |
| Machine-generated | 90,835 (about 76%) |
| Automated responses opposing restrictions | All but two |
| Support among valid respondents | 79.6% |
Impact
Short-term: The study puts pressure on governments to review how they screen online consultation responses for automated and AI-generated submissions.
Long-term: As AI-written text becomes harder to detect, consultations may need stronger verification, which could change how easily the public can take part in policymaking.
Who is affected: Governments and regulators that run public consultations, citizens and groups who take part in them, public health bodies, and industries whose regulation depends on consultation outcomes.
Key Takeaway
A Tobacco Control study found that 90,835 of 118,756 responses, about 76%, to the UK's 2023 disposable vapes consultation were bot-generated, and almost all opposed restrictions.
Questions and Answers
How many responses to the UK disposable vapes consultation were made by bots?
According to a study in Tobacco Control reported on October 28-29, 2026, 90,835 of the 118,756 submissions, about 76%, were machine-generated. Almost all of them opposed restrictions on disposable vapes.
Did the bot responses change the outcome?
No. The Department of Health and Social Care flagged and excluded them. Among valid respondents, 79.6% supported restrictions, and the UK went ahead with a ban on the sale and supply of disposable vapes in England.
Who was behind the automated responses?
The researchers said the volume and consistency of the responses suggest a targeted campaign, but that it is not possible to identify who was responsible.
How were the automated responses detected?
Officials noticed entries arriving very frequently from a small cluster of internet addresses, with heavily duplicated wording, and used bot-detection screening to identify them.
Disclaimer: Prepared by the Peepals newsroom from publicly available sources with AI assistance. Information is accurate to the best of our knowledge at the time of publication and may change. Images may be representative. Not professional advice. Report an error via our contact page.
By Anika Sharma, AI & Innovations EditorCovers AI & Innovation and more for Peepals Tech. All stories by Anika Sharma
UK Home Office embedding AI in immigration with little openness, report says
Microsoft, Alphabet results: Azure up 48%, Google Cloud doubles on AI demand
Get the day's stories in your inbox
Peepals Weekly: the week decoded, free. One email a week, unsubscribe in one click.
Anika Sharma is a pen name of the Peepals Global Editorial Team, which researches, fact-checks and publishes every story. The portrait is an illustration.
