The probe will assess whether X has properly evaluated and mitigated the risks associated with deploying its AI tool Grok on the platform within the EU. In particular, the investigation focuses on the spread of illegal content, including manipulated sexually explicit images, potentially involving depictions of child sexual abuse.

The background: According to The New York Times and the Center for Countering Digital Hate (CCDH), Elon Musk’s AI chatbot Grok generated and published at least 1.8 million sexualized images of women on X within just nine days. Roughly 65% of the images contained sexualized depictions of men, women, or children, CCDH reported.

The surge began on December 31, after Musk shared a bikini image of himself generated by Grok. Users then began prompting the chatbot to digitally undress real photos of women and children. X restricted this functionality only on January 8.

If violations are confirmed, X could face further penalties. In December 2025, the Commission had already fined X €120 million for misleading interface design, insufficient advertising transparency, and inadequate data access for researchers.

EU Commission Investigates Transition to Grok-Based Recommendation Algorithm

In addition, an ongoing investigation launched in December 2023 has been expanded. That case examines the effectiveness of X’s reporting and takedown mechanisms, its measures against illegal content such as terrorist material, and risks posed by its recommendation systems. The probe now also includes the recently announced shift to a Grok-based recommendation algorithm.

The investigation will assess the extent to which the significant bias observed in newer Grok systems is reflected in the so-called Phoenix Scorer. Researchers have already demonstrated in 2022 how quickly even relatively simple recommendation algorithms can become manipulative.