Most Dutch EU tobacco rule submissions were generated by a tobacco company’s AI tool

2026-07-05 · Source: Welcome to the Artificial Intelligence Incident Database · Field: Government & Public Sector — Public Policy & Governance, Regulatory & Compliance · Depth: Fundamental Awareness, quick

Summary

An investigation by NOS and Pointer revealed that nearly three-quarters of Dutch submissions to a European Union consultation on stricter tobacco rules were fully written by artificial intelligence. Most of these 786 responses, specifically 71%, were traced to Philip Morris International's (PMI) "Your Voice Your Choice" AI tool, which has since been taken offline. Across 15 EU countries, researchers analyzed over 65,000 responses, finding 35% were highly likely AI-generated, with the Netherlands ranking second. The PMI tool, accessed via QR codes, guided users to generate first-person responses opposing stricter regulation, often including arguments not explicitly selected. This campaign has prompted "very concerned" reactions from the European Commission and triggered political and regulatory actions in the Netherlands, despite PMI's claims of voluntary participation and improved clarity.

Key takeaway

For policy makers evaluating public consultation feedback, you must implement robust AI detection and source verification protocols. The demonstrated use of AI tools by industry to generate and submit biased responses, as seen with Philip Morris International, risks misrepresenting genuine public sentiment. You should scrutinize high-volume, similarly worded submissions to ensure democratic processes remain unmanipulated by automated influence campaigns.

Key insights

Corporate AI tools can subtly manipulate public consultations, generating biased responses at scale.

Principles

• AI-generated content can mimic genuine public opinion.
• Guided response systems can steer user input.
• Industry-backed campaigns face regulatory scrutiny.

Method

A guided AI tool presented users with questions, then generated first-person responses opposing regulation, often adding unselected arguments for submission.

In practice

• Implement AI detection for public consultation submissions.
• Scrutinize high-volume, similar-phrased public inputs.
• Monitor industry-sponsored "citizen engagement" tools.

Topics

• AI-generated content
• Public consultations
• Tobacco regulation
• Lobbying tactics
• Philip Morris International
• Influence campaigns

Best for: Executive, CTO, VP of Engineering/Data, Policy Maker, AI Ethicist, Legal Professional

Related on AIssential

• Philip Morris rejects press claims it ‘rigged’ the EU tobacco consultation with AI — AI-assisted public consultation responses, particularly from corporate-backed platforms, raise concerns about democratic process integrity.
• Philip Morris used AI to generate public opposition to EU rules — Philip Morris used AI to generate fake public comments, simulating grassroots opposition to EU vape regulations.
• Penguin v. OpenAI: AI-assisted production of market substitutes in the very formats that already plague children’s publishing (fake titles, lookalike covers, rapid self-publishing). — AI models reproducing copyrighted content on demand can constitute actionable copyright infringement.
• Q&A: Do AI and bogus respondents threaten polling’s future? — The rise of AI-generated and faked responses in polling necessitates a return to rigorous, probability-based human sampling for trustworthy public opinion data.
• AI Is Reshaping PR Forever: Here's How — AI fundamentally transforms PR by enabling data-driven communication, enhancing efficiency, and fostering creativity for superior campaign outcomes.
• How America Turned Against AI According to the Poll Data: A (Very Big) Compilation — Poll data consistently shows a rapid, bipartisan collapse in American public trust for AI, its infrastructure, and industry leaders.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Welcome to the Artificial Intelligence Incident Database.