MEPs Vote for €35M Fines on Social Media for Non-Consensual AI Images
The European Parliament approved new rules allowing fines up to €35 million on social media firms for AI-generated non-consensual sexual images. The ban, effective by year-end, aims to protect individuals, primarily women, from humiliation. The European Council will formally approve the rules on June 29, with enforcement starting December 2.
The European Parliament in Strasbourg has voted 423 to 57, with 174 abstentions, to approve new rules imposing fines on social media firms that allow their AI tools to create non-consensual sexual images. These rules, part of the Digital/AI omnibus bill, specifically prohibit making available systems used for such images of men, women, and children.
Firms ignoring this after December 2026 will face fines up to €35 million or 7% of their total worldwide annual turnover. The European Council will formally approve the rules on June 29, with enforcement beginning December 2. Michael McNamara, MEP for Ireland South and co-rapporteur for the Civil Liberties, Justice and Home Affairs committee, highlighted the ban on AI nudification apps, stating they impact real people, predominantly women, for humiliation and degradation. He expressed pride in the parliament's fight for this ban, which will take effect by year-end.
Other measures in the bill include mandatory watermarks for AI-generated content, an EU ban on systems creating non-consensual explicit images unless they have adequate safeguards, a clearer definition of AI «safety components», and extended AI tool exemptions for some small mid-cap enterprises.