STOP warns of the tobacco industry's use of artificial intelligence to influence the discourse on the subject

According to STOP, a global network of academic and public health organizations that monitors tobacco industry practices, there is evidence that the industry seeks to gain public trust and influence online opinion.[1]. These efforts appear in familiar spaces on the web, such as news sites, search platforms, and social networks.

But major companies in the sector seem to be going further, also acting in less visible ways to promote their activities and products. Research suggests that artificial intelligence tools are being used to spread disinformation online, promote narratives favorable to the industry, and possibly expand their product offerings.

How the industry seeks to influence AI-generated content

When a search engine or AI-powered assistant answers a question, it relies on what has already been published online, whether that information is accurate or not. Large companies therefore seek to influence these answers by multiplying content aligned with their brand image across various online spaces, such as social media. or YouTube, which are widely used sources of information for young people.

Furthermore, the influence strategy is reinforced by the mass of misleading content that has been present for decades, such as biased scientific studies funded or conducted by the industry, corporate communications, advertorials, and other industry-biased and shaped content. This work is often used to downplay the risks associated with tobacco products or to challenge regulatory policies, without its links to the industry always being clearly visible.

Even when content is factually accurate, it can still be biased by the selection of certain elements of industry discourse, the omission of risks, or the repeated emphasis on "harm reduction" arguments echoing the tobacco manufacturers' narrative. Researchers also believe the industry could strengthen its presence on Reddit, a platform often used as a source of content deemed authentic by AI models. While proving a link to the industry is difficult, researchers found content on Reddit promoting snus and nicotine pouches. The number of posts about nicotine pouches increased between 2019 and 2021, and more than half of them presented the pouches in a positive light.

The risk of organized disinformation

Researchers also warn against the potential use of automated bots or coordinated disinformation campaigns to disrupt AI tools used in the fight against smoking. Such content can distort social media monitoring programs, sentiment analysis tools, and also public consultations on tobacco-related laws.

In this context, research, public policy, and consumer decisions risk being influenced by biased information favoring industry. According to the consortium of organizations, these strategies may only be the beginning, and other uses of AI by the sector could yet emerge.

STOP also mentions the announced development of a major AI model specifically designed for the tobacco industry and consumers, Tobacco Titan. This type of tool could provide information on products, regulations, health and safety, as well as suggest new flavors, brands, or accessories.

Alongside essential citizen vigilance, AI could disseminate public health messages

Used wisely, artificial intelligence can also serve public health. Research suggests that well-configured chatbots can offer personalized interventions to help people quit smoking. Other tools, such as those that monitor social media like Canary, are already being used to document online tobacco marketing practices and report violations of advertising regulations.

STOP also cites an experimental AI developed by the Tobacco Control Research Group at the University of Bath, capable of assessing the accuracy of tobacco-related claims in real time. Compared to experts in the field, this tool proved highly accurate and faster, although further testing is still needed. These tools could help public health advocates track misinformation and better hold the industry accountable for its practices.

With the strengthening of protective advertising bans, the tobacco industry is particularly expanding its presence on digital channels, which are more difficult to control, and is thus able to circumvent and violate existing regulations. Without oversight, this digital presence could reach millions of young people and contribute to maintaining or even reigniting the smoking epidemic, given that three-quarters of the world's population uses the internet. Therefore, until AI is perfected, internet users' vigilance remains the best way to control the excesses of the tobacco and nicotine industry. In France, an organization like the Audiovisual and Digital Communication Regulatory Authority (Arcom) has decided to grant the National Committee Against Smoking (CNCT) the status of "trusted whistleblower" in order to strengthen vigilance and pursue the authors of content illegally promoting tobacco, vaping, or nicotine products.

©Generation Without Tobacco

AD

[1]Big Tobacco May Be Influencing AI-Generated Content Online, STOP, published on June 25, 2026, accessed on June 26, 2026

National Committee Against Smoking |