Skip to content
Submit incident
AII-1000-81171
AllegedThe record is a lawsuit, complaint or accusation that has not been decided.

Philip Morris AI Tool Allegedly Generated Over 65,000 Responses Against EU Tobacco Rules

Dutch broadcasters Pointer KRO-NCRV and NOS reportedly found that Philip Morris International deployed an AI tool called 'Your Voice, Your Choice' to generate over 65,000 responses to an EU public consultation on stricter tobacco rules, with most responses opposing tighter regulations. Researchers using AI detection software determined that approximately 35 percent of over 65,000 responses analyzed across 15 EU countries were highly likely fully AI-generated, and the European Commission expressed serious concern about industry-backed manipulation of the consultation process.

19 May 2026 Β· Philip Morris International Β· Misinformation and synthetic media
19 May 2026
AII-1000-811712 sources ↓
Quick facts
Category
Misinformation and synthetic media
Source
nltimes.nl
Date
19 May 2026
Drafted by
Raidu Incident Detection

What happened

Philip Morris International allegedly used its 'Your Voice, Your Choice' platform to direct users via QR codes on yellow posters in tobacco stores and on the IQOS website to an AI tool that generated responses to the European Commission's public consultation on stricter tobacco legislation. According to investigations by Dutch broadcasters Pointer KRO-NCRV and NOS, the system offered around 50 pre-defined responses, almost all opposing stricter regulations on flavourings, nicotine pouches, filters, and advertising.

Researchers analyzed over 65,000 responses across 15 EU countries using the AI detector Pangram and statistical models. They found that 35 percent were highly likely fully AI-generated, with 71 percent of the 786 Dutch responses examined assessed as almost certainly fully written by AI. In automated testing, researchers found that generated texts often included arguments not explicitly selected by users.

Reported impact

Over 80,000 responses were submitted to the European Commission's consultation, with one in three reported to have been generated by artificial intelligence. Of the 370 responses linked with high confidence to the PMI tool, only six supported tighter regulation, while the rest opposed it. The European Commission stated it is 'very concerned' about industry-backed campaigns to influence the results of the consultation.

Index analysis

Eleven of 12 selectable 'important aspects' on the tool produced arguments against stricter regulation, and all optional 'further concerns' were also negative toward regulation. The tool functioned as a guided response generator rather than a general chatbot, with participants answering questions about smoking status and policy preferences before the system produced first-person responses ready for submission.

Governance controls

Philip Morris stated that the platform 'uses an artificial intelligence tool to help users express their opinions' and that 'users remain fully in control: they can edit the output, and what they ultimately write...it is entirely up to them.' The European Commission responded that it condemns 'the alleged attempts by the tobacco industry to steer responses to the call for evidence' and stated 'It is crucial that citizens' views are not manipulated or misrepresented.' The tool has since been taken offline.

Classification

Category
Misinformation and synthetic media
Industry
Consumer products
Harm status
Actual harm
Intent
Intentional
Timing
Post-deployment
Harm types
Public interest, Reputational
Harmed
General public, Government
Failure modes
Manipulation or dark pattern
Consequences
Regulatory investigation
AI systems
Your Voice Your Choice
Regulations
GDPR (cited)

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

nltimes.nl
Philip Morris AI Tool Allegedly Generated Over 65,000 Responses Against EU Tobacco Rules
eunews.it
Philip Morris AI Tool Allegedly Generated Over 65,000 Responses Against EU Tobacco Rules