The recent controversy surrounding the sexual deepfakes generated by Grok, X’s artificial intelligence (see EUROPE 13779/14), continues to grow. X’s AI model has come under fire after introducing a ‘daring mode’ in the summer of 2025, which was quickly hijacked by users to generate non-consensual sexual images, sometimes involving underage girls.
Faced with these abuses, the European Commission confirmed on Thursday 8 January that it had ordered X to retain all internal documents and data relating to Grok’s functionalities until the end of 2026.
“We’re well aware that it’s not ‘daring’. It’s illegal, it’s revolting, it’s disgusting”, insisted Thomas Regnier, spokesperson for the institution. “Compliance with European law is not an option. It is an obligation”, he added.
In the European Parliament, the case is seen as a further test of the DSA’s credibility. The Renew Europe group has taken up the issue and on the same day, they sent a letter to the President of the European Commission, Ursula von der Leyen, calling for “swifter and more decisive action”, as well as the formal opening of an investigation against X in the context of the DSA.
In their letter, the MEPs consider that Grok’s actions constitute a serious “failure to comply with the risk mitigation and protection of minors obligations” as set out in the DSA. In view of the seriousness of the facts, the European Commission should “invoke Article 70 to order the platform to temporarily suspend or restrict xAI’s Grok chatbot for the duration of the investigation”, wrote the MEPs.
In addition to the European institutions, the case is also of concern to those involved in the fight against online sexual abuse. The Internet Watch Foundation, which acts as the UK hotline for reporting photos or videos of child sexual abuse on the Internet, has confirmed that it has discovered child pornography images of children aged between 11 and 13, generated via Grok and distributed on the Dark Web.
This case comes just one month after X was fined €120 million by the European Commission under the DSA on 5 December (see EUROPE 13767/1). At the time, the platform was found in breach for its lack of transparency regarding its algorithms, its lack of cooperation with researchers and the misleading design of its ‘blue checkmarks’.
This sanction added fuel to the fire in transatlantic relations, with the United States repeatedly claiming that European regulations are a disguised form of censorship against American companies. This new case involving X is likely to revive American criticism of the application of the DSA, which is deemed to be “abusive”. (Original version in French by Isalia Stieffatre)