The European Union opened a formal investigation into Elon Musk’s social media platform X after his artificial intelligence chatbot Grok produced nonconsensual sexualized deepfake images on the platform, the European Commission said. The EU executive also said regulators broadened an existing, separate inquiry into X’s recommendation systems after X indicated it would switch to Grok to help determine which posts users see.
In a statement, the European Commission said it was looking into whether X did enough, as required by the bloc’s digital regulations, to contain the risks of spreading illegal content, including “manipulated sexually explicit images.” The Commission also said the risks have “materialized,” exposing EU citizens to “serious harm.”
Brussels said the investigation would examine whether Grok is meeting obligations under the Digital Services Act, the bloc’s wide-ranging rulebook on keeping internet users safe from harmful content and products. The Commission said those risks include content that “may amount to child sexual abuse material.”
The scope of the probe, the Commission said, covers Grok’s service on X and not Grok’s website and standalone app. The Commission said that distinction reflects how the Digital Services Act applies to the bloc’s largest online platforms.
Henna Virkkunen, an executive vice president at the commission responsible for tech sovereignty, security and democracy, said non-consensual sexual deepfakes of women and children are “a violent, unacceptable form of degradation.” She said the investigation will determine whether X met its legal obligations under the DSA, or whether it treated the rights of European citizens—including those of women and children—as collateral damage of its service.
X responded by pointing to a Jan. 14 statement, according to an X spokeswoman who directed The Associated Press to that earlier message. In that statement, X said it remains “committed to making X a safe platform for everyone” and has “zero tolerance” for child sexual exploitation, nonconsensual nudity, and unwanted sexual content.
The Jan. 14 statement also said X would stop allowing users to depict people in “bikinis, underwear or other revealing attire,” but only in places where it has been deemed illegal, AP reported. In Brussels, regulators said Grok’s image-generation and editing capabilities helped spark backlash by enabling users to undress people in sexually explicit deepfakes, and that researchers said some images appeared to include children.
AP reported that xAI launched Grok’s image tool last summer, but that the issue escalated in late December when Grok seemingly granted many requests to modify images posted by others. The problem also spread more quickly, AP said, because Grok’s responses on X are publicly visible.
The EU said it would not set a deadline for resolving the case. The investigation could end with X pledging to change its behavior, or with a “hefty fine,” AP reported.
The new probe comes after Brussels fined X in December 120 million euros ($140 million at the time) as part of an earlier ongoing Digital Services Act case. In that earlier action, the Commission said shortcomings included blue checkmarks that broke rules on “deceptive design practices.”
The bloc has also been scrutinizing X over allegations that Grok generated antisemitic material and has asked the site for more information, AP reported. Outside the EU, AP said Malaysia and Indonesia blocked access to Grok earlier this month; Malaysian authorities later said they lifted a temporary restriction after X implemented additional security and preventive measures, while regulators said they would continue monitoring.
In the United States, AP reported that attorneys general in 35 states wrote to X last week urging it to disclose its plans to prevent Grok from creating nonconsensual sexualized deepfake images and to explain how it will eliminate existing content. The letter said, “We strongly urge you to be a leader in this space by further addressing the harms resulting from this technology.”