European Union lawmakers moved to ban so-called nudifier apps, systems that use Artificial Intelligence to alter images of real people into nude or sexually explicit versions without consent. The change was adopted by the European Parliament’s committees on the Internal Market and Consumer Protection and on Civil Liberties and Justice on 18 March with 101 votes in favour, 9 against, and 8 abstentions. The proposal forms part of the ongoing review of the Artificial Intelligence Act, the European Union’s main regulation on the use of Artificial Intelligence.
The push accelerated after controversy around Grok, the assistant on X, which was linked to the mass generation of content that virtually undressed real people without their knowledge. According to data released by the NGO Centre for Countering Digital Hate, GROK is said to have produced 3 million sexually explicit images and 20,000 artificial reproductions of child sexual abuse over an eleven-day period between late 2025 and early 2026, before the platform controlled by Elon Musk announced measures to combat their spread. Lawmakers framed the ban more broadly than a response to a single platform, presenting it as a measure that had been strongly demanded by the public.
The committees also approved wider amendments intended to make the Artificial Intelligence Act simpler and more flexible for businesses. They supported delaying some obligations for high-risk systems, including those using biometrics or operating in sectors such as infrastructure and healthcare. Current legislation requires companies to comply with the new rules by 2 August this year, but lawmakers argued that the definition of key standards is unlikely to be ready by that date. They proposed 2 December 2027 for high-risk Artificial Intelligence systems listed directly in the law, and 2 August 2028 for systems already covered by other European Union safety and market surveillance rules.
On transparency rules, lawmakers supported more time for compliance with watermarking requirements for Artificial Intelligence-generated content, but with a shorter delay than the European Commission had proposed. The committees suggested an extension until 20 November 2026 instead of 2 February 2027. Other changes include extending support measures beyond small and medium-sized enterprises to Small Mid-Cap Enterprises, easing obligations for products already regulated under sector-specific European laws, and allowing companies in limited cases and with safeguards to use personal data to identify and correct bias in Artificial Intelligence systems. The amendments will go to a plenary vote in Strasbourg on 26 March, after which negotiations with Member States in the Council of the European Union may begin.
