Government consultation on disposable vapes ban flooded by AI bots, study finds

Unbiased summary

A study published in the journal Tobacco Control found that 90,835 of 118,756 responses (76%) to the UK government's 2023 consultation on its tobacco and vapes bill were machine-generated. The Department of Health and Social Care used bot-detection methods to identify the automated submissions, which were often duplicated and sent at high frequency from a small number of IP addresses. All but two of the bot responses opposed proposed restrictions on disposable vapes. The valid responses showed 79.6% support for the restrictions. The researchers, Dr Nason Maani and Prof Anna Gilmore, said this appears to be the first documented instance of large-scale automated interference in a formal public health policy consultation in the UK. The government excluded the bot responses from its decision-making, and the ban on disposable vapes was subsequently enacted.

Generated with deepseek-flash (DeepSeek)

Coverage by outlet

The Guardian left

Angle Frames the story as a threat to health policymaking and democracy by powerful interests, emphasising the novelty of AI interference.

Bias Emphasises the 'failed attempt' to stop the ban and quotes an expert warning about future campaigns becoming harder to catch and powerful interests looking like a mass movement. It includes a quote from Action on Smoking and Health, which opposes vaping, and frames the interference as a threat to public health. It omits the researchers' caution that the origins of the campaign are not known and does not mention the specific support level among valid respondents.

AI bots flooded 2023 UK government consultation on disposable vapes ban

The Independent centre-left

Angle Presents a straightforward account of the study, highlighting the scale of bot interference and the reversal of public opinion once bots were excluded.

Bias Includes the key fact that after excluding bots, 79.6% of valid respondents supported restrictions, which is more balanced than some outlets. However, it omits the researchers' mention of prior tobacco industry efforts to manipulate consultations, which The Mirror includes. It also does not quote any experts beyond the researchers, and its headline uses 'influence outcome' rather than 'derail' or 'hijack'.

Huge swarm of AI bots flooded vape ban consultation to influence outcome, study finds

The Mirror centre-left

Angle Frames the story as a 'co-ordinated campaign' to 'manipulate' results, explicitly linking it to tobacco industry efforts.

Bias Emphasises the researchers' statement that the scale and consistency suggest a coordinated campaign consistent with prior tobacco industry manipulation. It includes a quote about transnational tobacco companies seeking to swamp consultations, which is not highlighted in some other outlets. It omits the fact that the researchers said it is not possible to ascertain the commercial origins, and it does not mention that the bot responses were excluded from decision-making until later in the article.

AI robots 'hijacked Government consultation on laws to curb vape sales'

GB News right

Angle Focuses on the bot interference as a fraudulent attempt to mislead, but also provides detailed factual context about the legislation and study methodology.

Bias Uses terms like 'bogus submissions' and 'fraudulent responses' but does not explicitly frame it as a tobacco industry plot. It includes the detail that researchers could not establish who was responsible, which is more cautious than some. However, it does not mention the researchers' reference to prior tobacco industry manipulation, and its headline emphasises the 75% figure without noting the support level among valid responses.

New study reveals 75% of respondents to Government's smoking and vaping laws consultation were bots