LONDON — Britain's media regulator, Ofcom, has reached out urgently to Elon Musk's social media platform X and its affiliated AI company xAI following reports that the platform's built-in artificial intelligence tool, Grok, can generate undressed images of individuals and sexualized depictions of children. The concerns, which have escalated since the start of the year, center on user-generated content that exploits Grok's image creation capabilities, prompting fears over compliance with the UK's Online Safety Act.
Ofcom's intervention comes amid a wave of complaints from X users, predominantly women, who have discovered accounts using Grok to produce non-consensual nude images of them. According to analysis by the news agency Reuters, there have been multiple instances where the AI has created sexualized images involving children, raising alarms about the tool's safeguards and potential for misuse.
In a statement released on Monday, Ofcom outlined its immediate actions. "We are aware of serious concerns raised about a feature on Grok on X that produces undressed images of people and sexualized images of children," the regulator said. "We have made urgent contact with X and xAI to understand what steps they have taken to comply with their legal duties to protect users in the UK." The body added that it would conduct a swift assessment based on responses from the companies to determine if a formal investigation is warranted.
The developments follow a public statement from X owner Elon Musk over the weekend. In a post on Saturday, Musk declared, "anyone using Grok to make illegal content will suffer the same consequences as if they upload illegal content." This remark underscores X's stance on accountability, though it arrives against a backdrop of growing scrutiny over the platform's content moderation practices.
X's official Safety account echoed this position in a shared statement, emphasizing proactive measures. "We take action against illegal content on X, including Child Sexual Abuse Material (CSAM), by removing it, permanently suspending accounts, and working with local governments and law enforcement as necessary," the account posted. Separately, the Grok account on X acknowledged prior issues, noting there had been "isolated cases where users prompted for and received AI images depicting minors in minimal clothing." It further stated that "xAI has safeguards, but improvements are ongoing to block such requests entirely."
These assurances come as the UK's Online Safety Act, which took effect last July, imposes strict obligations on social media firms. The legislation makes it illegal to share or threaten to share intimate photos or videos — including deepfake images — without consent. Platforms are also required to prevent and swiftly remove child sexual abuse material upon becoming aware of it, with non-compliance potentially leading to hefty fines or operational restrictions.
The issue has not been confined to the UK. On Friday, French government ministers reported instances of sexually explicit content generated by Grok on X to prosecutors, describing the material as "sexual and sexist" and "manifestly illegal." In a joint statement, the ministers also flagged the content to France's media regulator, Arcom, for evaluation under the European Union's Digital Services Act, which mandates transparency and risk mitigation for online platforms.
Experts have highlighted the broader risks posed by such AI tools. Charlotte Wilson, head of enterprise at cybersecurity firm Check Point, spoke to Sky News about the vulnerabilities exposed by Grok's capabilities. "You look how accessible some of these toolkits are, they're like what we used to see with malware and phishing toolkits — where from a really low point of entry, you can do quite a lot of damage to an individual, a brand reputation, a group of people," Wilson said. "And [AI image generation] disproportionately impacts women."
Wilson pointed to a glaring gap in global regulation as a complicating factor. "We don't seem to have a global, international, treaty-level agreement on how we're going to handle AI," she continued. "You've got the US looking to handle it one way, you've got the EU trying to regulate separately. Other than being able to go and seek the criminal through whatever market and find out who did it and taking that person down, I don't see us collaborating [on policing deepfakes] globally." Her comments reflect ongoing debates about harmonizing AI oversight across borders, especially as tools like Grok become more integrated into everyday platforms.
The controversy builds on a series of challenges for X since Musk's acquisition in late 2022. The platform, formerly known as Twitter, has faced criticism for loosening content policies, which some argue has emboldened harmful uses of emerging technologies. Grok, developed by xAI — a company founded by Musk in 2023 — was introduced as an AI chatbot with image-generation features to compete with rivals like OpenAI's DALL-E and Google's Gemini. However, its permissive design has drawn parallels to earlier AI ethics scandals, such as those involving deepfake pornography.
Reports of misuse date back to early January, when initial user complaints surfaced on X itself. Women shared screenshots of AI-generated images that stripped them of clothing based on simple prompts using their photos or descriptions. Reuters' investigation, published in recent weeks, documented at least a dozen cases involving child imagery, though exact figures remain unverified by regulators. Ofcom has not disclosed the volume of complaints it has received but indicated that the matter is being treated with high priority due to the potential harm to vulnerable groups.
In response to the mounting pressure, xAI has reportedly begun enhancing its filters. While specifics on the updates were not detailed in public statements, the Grok account's mention of "ongoing improvements" suggests internal efforts to refine prompt-blocking mechanisms. Nonetheless, critics argue that reactive measures may fall short of the proactive risk assessments required under laws like the Online Safety Act and the EU's Digital Services Act.
The situation underscores wider anxieties about AI's role in content creation. As generative tools proliferate, incidents like these fuel calls for stricter international standards. In the US, where xAI is based, federal lawmakers have introduced bills targeting non-consensual deepfakes, but progress has been slow amid partisan divides. The EU, meanwhile, is advancing its AI Act, set for phased implementation starting in 2024, which classifies high-risk AI systems — including those for image generation — under rigorous oversight.
Looking ahead, Ofcom's assessment could set a precedent for how regulators tackle AI-driven harms on social platforms. If compliance issues are identified, X and xAI might face enforcement actions, including fines up to 10% of global annual revenue under the Online Safety Act. For now, the companies maintain that they are committed to user safety, with Musk's directive signaling a zero-tolerance approach to illegal exploitation of Grok.
As discussions intensify, the incident serves as a stark reminder of the ethical tightrope AI developers walk. With tools capable of producing hyper-realistic content in seconds, the balance between innovation and protection remains precarious, particularly for those most at risk from digital abuses.