The European Parliament voted Thursday to prohibit AI systems that generate explicit sexual imagery without consent and postponed regulations for high-risk AI systems.
European Parliament Approves AI Restrictions
The European Parliament has adopted its position on the Digital Omnibus on AI, including a package of amendments. These amendments delay the implementation of requirements for high-risk AI systems.
The Parliament also approved an amendment banning the use of artificial intelligence systems used for “nudification” – those that digitally undress people using AI.
Ban on ‘Nudifier’ Applications
The approved amendment introduces a ban on systems categorized as “nudifiers,” which utilize AI to digitally remove clothing from individuals.
Scope of the Ban
According to a statement from the European Parliament, the ban will apply to systems that “use artificial intelligence to create or modify images of an explicitly sexual or intimate nature, depicting a recognizable, real person without their consent.”
Lawmaker Commentary
During the session, Member of Parliament Michael McNamara (Renew) stated that “artificial intelligence that creates images without consent will be prohibited. Sexualizing someone’s image using AI will be impossible in Europe, nor will it be possible to steal the image of public figures.”
New Timelines for AI Obligations
Requirements for high-risk AI systems will now be applicable from December 2, 2027. This includes systems utilizing biometric data and those used in critical infrastructure, education, employment, essential services, law enforcement, justice administration, and border management.
For AI systems covered by EU sectoral regulations concerning safety and market surveillance, the proposed deadline is August 2, 2028.
Suppliers have until November 2, 2026, to comply with regulations regarding the inclusion of watermarks on audio, graphic, video, or text content generated by artificial intelligence to indicate its origin.
AI and Small and Medium-Sized Enterprises
Members of the European Parliament support allowing service providers to process personal data to detect and correct errors in AI systems, but with safeguards ensuring this occurs only when absolutely necessary.
Regulatory Overlap
Amendments stipulate that obligations under the AI Act may be less stringent for products already regulated by sectoral legislation, such as medical devices, radio equipment, and toy safety, to prevent duplication of EU regulations.
Lawmaker Perspective on Bureaucracy
Arba Kokalari (EPP) commented that “today’s vote is a small step for artificial intelligence and Europe, but a big step towards cutting bureaucracy.”
Part of Broader Digital Package
The adopted regulations are part of the seventh Omnibus package concerning simplification and digital rules. The European Parliament is also working on changes to data usage and protection regulations and the introduction of a European business portfolio.