A UK MP is requesting the High Court to compel xAI to halt Grok from producing sexualized images of her.

      Jess Asato, the Labour MP representing Lowestoft, has requested the High Court in London to compel Elon Musk's xAI to prevent its Grok chatbot from generating sexualized images of her. Her legal team argues that this case could establish the extent to which AI companies are accountable for the content their tools produce.

      Asato is pursuing a court order mandating xAI to implement what her filing labels as “effective and permanent technical measures” to stop Grok from creating manipulated images of her, according to a Reuters report from July 28.

      This request builds upon a prior claim she filed in June, alleging misuse of private information and violations of UK data protection laws, and seeking damages alongside an official acknowledgment of the company's unlawful actions.

      The case arose after Grok was prompted by users to create images earlier this year, subsequent to Asato publicly criticizing Musk and the chatbot. Among the evidence referenced in her claim is a video in which she contends she is depicted being chloroformed and prepared for sexual assault.

      Asato has expressed a desire to rebalance individual rights against the immense power of major tech firms, with her lawyers arguing that the construction and training of Grok facilitated the production of such content.

      This argument is central to the dispute. Her solicitor, Ravi Naik, stated that the design decisions by the company “should carry legal consequences,” and that the remedy sought by his client “includes a court order to ensure compliance if xAI refuses to do so.”

      The claim highlights internal directives that, according to Asato, instructed the model to decline assistance with criminal activities while indicating that there were “no restrictions on adult sexual content or offensive content.” xAI has not yet submitted a defense, and none of the allegations have been adjudicated in court.

      The company has not responded to inquiries regarding the claim, and Musk has not publicly addressed it, a silence that has become familiar.

      Earlier this year, Apple privately threatened to remove Grok from its App Store over non-consensual deepfake images, and prosecutors in Paris initiated an investigation into sexual imagery of children related to the tool, though the company has largely refrained from engaging with both situations.

      The British context complicates matters for Musk, who has criticized the UK's Online Safety Act as one of the strictest in the world and a potential threat to free speech. This law, which has been in effect since 2023 and is overseen by Ofcom, requires platforms to address illegal content, and the UK has also taken steps to criminalize the creation of non-consensual sexual deepfakes, rather than just their distribution.

      Asato's claim is civil in nature and focuses less on content moderation and more on whether a company can be directly held accountable for the intended functionality of its system.

      Image-based abuse has rapidly evolved within British law, much of which was established prior to the advent of generative tools capable of producing material at scale.

      Grok has faced similar scrutiny before, not just in Britain. In March, EU lawmakers agreed to incorporate an explicit ban on non-consensual intimate deepfakes into the bloc’s AI Act, a move partly prompted by an earlier scandal involving the tool.

      The pattern observed in Paris, Brussels, and now London involves regulators and judges seeking to apply pre-existing legal frameworks to assess their relevance in the context of new technologies.

      What Asato is seeking is more specific and actionable than a policy; it is a court order, enforceable by a judge, requiring xAI to alter Grok’s functionality or face consequences for non-compliance.

      Whether the High Court grants this request, and under what conditions, will significantly determine who bears the risk when a chatbot generates content that its subject has not consented to. A hearing date has yet to be established.