Parlement européen — questions et conflits d'intérêts

« Question for written answer E-003332/2026 to the Commission Rule 144 Sandro Gozi (Renew) Subject: ChatGPT advertisements across EU countries: transparency, profiling and consumer protection Since 24 August 2026, OpenAI has begun showing advertisements in ChatGPT across 31 European countries. These advertisements have reportedly been selected based on conversation context and displayed alongside AI-generated responses. OpenAI states that advertisers cannot access conversations, ads are clearly labelled and separate from responses, and advertising does not influence answers. This raises concerns about transparency, profiling, and consumer protection, particularly where commercial content may be selected using sensitive information disclosed during a conversation, and where the conversational format may blur the line between AI recommendations and paid promotion. Can the Commission clarify: 1. Whether it has assessed this development’s compatibility with relevant provisions of the General Data Protection Regulation1, the Digital Services Act2, the AI Act3, and the Unfair Commercial Practices Directive4? 2. Whether the existing EU legal framework ensures users can clearly distinguish AI output from paid content, and prevents targeting based on sensitive information disclosed during AI interactions or exploitation of user trust? 3. Whether, if gaps in the legal framework exist, it will take measures to address the risks from such practices, including the designation of ChatGPT as a very large online search engine (VLOSE)? Submitted: 25.8.2026 »

Lire le texte entier →