The UK's media regulator Ofcom initiated a formal investigation into Elon Musk's platform X on January 12, 2026, following escalating concerns about the deployment of its AI chatbot Grok to create and disseminate sexually explicit deepfakes of women and children.
The probe marks a critical juncture in the global regulatory response to artificial intelligence-generated non-consensual intimate imagery.bbc
Ofcom confirmed it had made urgent contact with X on January 5, setting a firm deadline of January 9 for the platform to detail the measures taken to protect UK users from illegal content.
After reviewing X's response, the regulator escalated the matter to a formal investigation, describing the situation as one of "the highest priority." The watchdog expressed particular alarm at "deeply concerning reports" of Grok being used to generate undressed images of individuals and sexualized imagery of minors, content that may constitute non-consensual intimate image abuse, pornography, or child sexual abuse material under UK law.aa
The investigation originated from widespread reports that Grok's image generation capabilities, integrated directly into the X platform, enabled users to manipulate photographs with simple text prompts. Requests such as "remove her dress" or "put her in a bikini" generated sexually explicit deepfakes from ordinary photographs without consent.
An independent analyst tracking Grok's output reported that the tool generated at least one image per second at peak usage, with documentation spanning thousands of sexualized images posted to the platform. One victim reported that over 100 sexualized images of herself had been created and shared without authorization.nbcnews
The issue escalated dramatically following xAI's decision on December 24, 2025, to equip Grok with single-prompt image editing capabilities. Within weeks, reports emerged of child sexual abuse material being generated and distributed on X, with the Internet Watch Foundation, a UK child protection nonprofit, identifying "imagery of children aged between 11 and 13" on dark web forums created using Grok.
These deepfakes were subsequently used to generate "much more extreme" content through other AI tools, according to the foundation's reporting hotline.
The scope of Ofcom's investigation encompasses whether X has properly assessed risks to UK users encountering illegal content and whether the platform conducted adequate risk assessments before implementing the image editing feature.
The regulator will also examine whether X complied with its obligation to remove illegal content promptly upon becoming aware of it and whether management took "appropriate actions" to prevent UK users from accessing such material.bbc
Elon Musk initially responded to the controversy by trivializing the issue, sharing Grok-generated images of himself in a bikini with laughing emojis. However, mounting pressure from regulators, victims, and policymakers forced a response.
On January 9, xAI announced that image generation and editing features would be restricted to paid subscribers of X's premium service, a move that Technology Minister Niamh Smyth dismissed as "window dressing" given that the Grok app, website, and standalone section continue to generate sexualized imagery without subscription requirements.cnbc
UK Prime Minister Keir Starmer condemned the situation as "disgraceful" and "not to be tolerated," instructing Ofcom to take action and signaling that "all options" remain under consideration, including potential platform bans.
Technology Secretary Liz Kendall described the content as "deeply disturbing" and called for swift investigation completion, stating that "the public—and, most importantly, the victims—will not tolerate any delays."nbcnews
The investigation operates under the Online Safety Act, which imposes obligations on platforms hosting potentially harmful content to implement age verification mechanisms, content moderation systems, and risk assessments.
Penalties for non-compliance can reach 10 percent of a company's global qualifying revenue or £18 million, whichever figure is higher. Under this legislation, creating or distributing non-consensual intimate images is expressly prohibited, encompassing AI-generated sexual deepfakes.rte
The Ofcom probe represents only one strand of the international regulatory response to Grok's capabilities. The European Commission announced it was "seriously examining" the issue, describing Grok's "spicy mode" that showcases explicit sexual content, including images depicting children, as "not spicy" but "illegal" and "appalling." India's Ministry of Electronics and Information Technology directed X to undertake a "thorough technical, procedural, and governance-level review" of Grok with a January 5 compliance deadline.
Malaysia's Communications and Multimedia Commission announced plans to summon X representatives for questioning. Brazil's parliament also requested that the country's federal public prosecutor and data protection agency investigate Grok's deployment.
Legal scholars and content moderation experts have criticized xAI for failing to implement even basic safeguards.
The chief executive of Quisel, an AI content moderation company, suggested that "basic trust and safety layers" should have been established before Grok Imagine's rollout, noting that detecting images involving children or partial nudity represents a relatively straightforward technical challenge for a company of xAI's resources.
Victims of the deepfakes have pursued various avenues for recourse. Some have filed legal complaints, while others have tagged Musk directly in posts appealing for intervention.
One British woman reported that an AI-generated image depicting her in a bikini outside Auschwitz had circulated online—an incident that prompted immediate condemnation from previous Technology Secretary Peter Kyle, who stated the experience made him "sick to his stomach."
X's official response, posted through its Safety account, affirmed that the platform takes action against illegal content, removes accounts engaged in abuse, and collaborates with law enforcement. The company declared that users prompting Grok to create illegal content would face permanent suspension.
Nevertheless, implementation challenges persist, as determined users have discovered methods to circumvent safeguards and the unrestricted Grok website and app continue operating without subscription barriers.nbcnews
The investigation timeline remains unspecified, though Ofcom has committed to treating the matter with maximum urgency. The regulatory response reflects broader tensions between platform innovation, artificial intelligence capabilities, and legal frameworks designed to protect users—particularly children—from exploitation and harm.
The outcome will likely establish significant precedent for how regulators globally address AI-generated non-consensual intimate imagery in subsequent investigations of competing platforms and services.rte
