UK Prime Minister Keir Starmer has issued a stark warning to Elon Musk’s X platform, declaring that the social media giant could lose its “right to self regulate” if it continues to fail in controlling its AI chatbot Grok, which has been generating sexualized deepfakes of women and children across its platform. The escalating government response marks one of the most significant regulatory confrontations between a major government and a major technology platform, with multiple enforcement mechanisms now in motion.
Starmer’s warning came as Britain’s media regulator Ofcom announced a formal investigation into X on January 12, examining whether the platform has violated the Online Safety Act. “If X cannot control Grok, we will—and we’ll do it fast because if you profit from harm and abuse, you lose the right to self regulate,” Starmer told Labour MPs on Monday. The Prime Minister characterized the deepfakes as “absolutely disgusting and shameful” and accused X of prioritizing “abusive users” over the women and children whose images have been exploited.
Grok has generated a flood of nonconsensual intimate images since late December, with users prompted the chatbot to remove clothing from real photographs of women and children or place them in sexually suggestive contexts using simple text commands. Ofcom described the reports as “deeply concerning,” noting that the undressed images “may amount to intimate image abuse or pornography—and sexualized images of children…may amount to child sexual abuse material.” At least one woman reported having over 100 sexualized images created of her without consent.
The regulatory response carries significant teeth. Ofcom can impose fines of up to £18 million or 10 percent of X’s worldwide revenue—whichever is greater. More drastically, if X fails to comply with regulatory orders, Ofcom can seek a court order forcing internet service providers to block access to the platform entirely in the UK. Under rare circumstances involving significant harm, courts may impose such “business disruption orders” immediately rather than as a last resort.
Technology Secretary Liz Kendall has urged Ofcom to move quickly, asking the regulator to set a clear timeline “as soon as possible” and not allow the investigation to drag on for “months and months.” The government is simultaneously fast-tracking enforcement of existing laws: starting this week, it is bringing into force legislation from the Data (Use and Access) Act, passed in June 2025, which makes creating nonconsensual intimate images a criminal offense. Additionally, the government plans to introduce new legislation explicitly making it illegal to supply online tools designed to create such images.
X’s response has proven inadequate in the eyes of regulators and government officials. The platform restricted Grok’s image generation feature to paying subscribers after international backlash, but this move drew scathing criticism. Starmer’s office called the paywall “insulting” to victims, describing it as merely converting an illegal feature into a premium offering rather than solving the underlying problem. Some non-paying users have reportedly continued generating sexualized images despite the restriction.
The international dimensions of the crisis have intensified regulatory responses. Indonesia and Malaysia both temporarily suspended access to Grok over the weekend, with Malaysia’s communications regulator criticizing xAI for relying on insufficient user-initiated reporting mechanisms rather than implementing robust technical safeguards. Meanwhile, the contrast between UK and US responses is stark: as British officials escalate enforcement, US Defense Secretary Pete Hegseth announced that Grok would be integrated into Pentagon networks alongside Google’s generative AI systems.
Elon Musk has dismissed the regulatory scrutiny as government censorship, claiming the UK wants “any excuse for censorship” and questioning why other AI services aren’t under similar investigation. He has characterized the controversy as beneficial, claiming it has increased X app downloads, and has attacked Starmer personally with his own AI-generated provocative image.
The investigation represents a pivotal moment testing the authority of Britain’s Online Safety Act, implemented in July 2024. This will be the most significant enforcement action to date, with potential consequences extending far beyond financial penalties if X is found noncompliant.
