EU Parliament and Dutch Court Ban Sexualized AI Deepfakes

The European Parliament approved a ban on artificial intelligence systems that generate sexualized deepfakes, specifically targeting nudifier apps that create intimate images of real people without consent. The legislative move follows widespread outrage over content produced by the Grok chatbot and includes a proposal to delay the implementation of rules for high-risk AI models until as late as August 2028. While the Parliament and EU member states have approved the proposals, they must still negotiate the final text.

Simultaneously, the Amsterdam Court issued a preliminary injunction ordering xAI and its Grok chatbot to stop generating and distributing non-consensual sexual imagery in the Netherlands. The ruling prohibits images that undress adults or children or place them in sexualized poses without permission. xAI faces daily fines of 100,000 euros for non-compliance and must cease offering Grok on the X platform if it breaches the order.

Legal action in the Netherlands was initiated by the non-profit group Offlimits after a March 9 demonstration showed Grok could still generate such imagery despite claimed safeguards. This judicial action coincides with a formal probe by the European Commission into X and a broader EU investigation into the rollout risks of the chatbot. Although X announced plans in January to stop the creation of sexualized deepfakes of women and children, regulatory and legal pressure continues to mount across the region.