UK media regulator says concerns persist over X's handling of illegal content

Britain's media regulator Ofcom said on Wednesday it remained concerned about Elon Musk-owned X's handling of illegal hate and terrorism-related content despite improvements in response times, while disclosing new restrictions on the Grok ‌AI chatbot. X met targets agreed with the regulator for reviewing reports of suspect content more quickly, according to the company's first quarterly performance report submitted to Ofcom.

However, the regulator said in a statement on Wednesday it had "ongoing concerns around the effectiveness of X’s response to taking down illegal hate and terror ‌content". The regulator said it would continue examining the data submitted by X and compare it with its own testing, adding that the company would ‌need to maintain its current performance.

X and Musk's SpaceXAI did not immediately respond to a request for comment. FOCUS ON COMPLIANCE AND REPORTING SYSTEMS

Separately, Ofcom also said its investigation into X had moved into a new phase in which it would decide whether the platform had met its legal duties to protect people in Britain from illegal content. Ofcom opened its investigation ⁠into X ​in January after reports that the Grok ⁠AI chatbot was being used to create and share sexualised images that could amount to illegal content, including deepfakes involving real people and children. Reuters later found Grok was still generating ⁠sexualised images of people when prompted, even where users explicitly said the subjects had not consented.

Ofcom said it found that some users who had not verified their age ​could still access pornographic content on X. It said the company had since changed its systems to ensure such material was blocked consistently ⁠from under-18s and unverified users. As a result, Ofcom said it would focus the remainder of its investigation on X's handling of illegal content and the platform's systems for reporting and dealing with ⁠complaints, while ​continuing to monitor its age-check measures.

If Ofcom has reasonable grounds to believe X failed to meet its duties, it said it would issue a provisional view before making a final decision. NEW RESTRICTIONS FOR GROK USERS IN BRITAIN

Ofcom said SpaceXAI, formerly xAI, had introduced further restrictions to Grok ⁠following discussions with the regulator. Users in Britain are now blocked from viewing, generating or sharing pornographic content through Grok's image-generation tools, according to Ofcom. Companion ⁠chatbots, such as "Ani", have also been barred ⁠from engaging in sexually explicit conversations.

The regulator said it would continue monitoring those changes and would keep engaging with SpaceXAI, having previously said gaps in the Online Safety Act limited its ability to investigate some activity within standalone ‌chatbot services.