EU Intensifies Enforcement Against Grok AI Over Illegal Deepfake Content

BRUSSELS: The European Commission has escalated regulatory action against Elon Musk’s Grok AI chatbot following widespread reports that the platform enabled users to generate sexualized deepfake images, including content depicting minors.

In January 2026, the Commission extended a data retention order requiring X to preserve all internal documents and data related to Grok until the end of 2026, utilizing enforcement powers granted under the Digital Services Act. The directive followed a scandal that erupted in late December 2025 when Grok’s new image-editing feature allowed users to create non-consensual intimate images.

Commission spokesperson Thomas Regnier condemned the content in stark terms, describing it as explicit sexual material with childlike images. He emphasized that such content is illegal, appalling, and has no place in Europe. The retention order ensures regulators can access evidence while assessing X’s compliance with DSA requirements.

The controversy centers on Grok’s “spicy mode” feature introduced in late December. Analysis by Paris-based NGO AI Forensics examined approximately 50,000 user requests and over 20,000 generated images between December 25, 2025 and January 1, 2026. The research found that 53 percent of analyzed images depicted individuals in minimal attire, with 81 percent classified as women. Around 2 percent appeared to show persons estimated at 18 or younger.

France’s Paris Prosecutor’s Office launched a criminal investigation into potential child pornography dissemination, while regulatory bodies in the UK, India, and Malaysia initiated separate probes. UK Prime Minister Keir Starmer condemned the images as disgusting and unlawful, with Business Secretary Peter Kyle warning that Grok could face a ban if accountability measures prove insufficient.

Under the DSA, very large online platforms like X must implement adequate risk-mitigation systems protecting against illegal content. EU Tech Sovereignty Commissioner Henna Virkkunen warned that failure to implement effective measures could trigger full enforcement, including fines up to 6 percent of global turnover or temporary platform suspensions.

X restricted Grok’s image generation features to paying subscribers on January 9, though reports indicated similar functionality remained accessible through alternative routes, prompting criticism that the measures were inadequate.