A lawsuit involving a UK MP may determine if xAI is held accountable for the outputs generated by Grok.

      The case revolves around the issue of authorship. When a user inputs a prompt and Grok produces a sexualized image of a real woman, who is responsible—the user at the keyboard or the company that created the tool?

      On June 3, Labour MP Jess Asato submitted a claim to the High Court in England, asking the court to determine whether an AI developer can be held directly liable for the outputs generated by its tool, marking the first significant UK examination of this issue.

      The details of the claim are particularly troubling. Asato, representing Lowestoft, states that users prompted Grok in January to create images of her in a bikini, alongside an explicit video described in her filing as depicting her being chloroformed and prepared for a sexual assault. These images were subsequently reshared and discussed on X, the platform owned by xAI, and used to generate additional content. Asato announced the claim herself, stating that she is “just one of thousands of women and even children” targeted by abusive AI deepfakes.

      The significance of the suit lies in its legal argument rather than its specific details. Asato is suing xAI for misuse of private information and violations of data protection law, arguing that the company is liable even if individual users submitted the prompts.

      She is seeking damages, a formal acknowledgment of the unlawful nature of her experience, and an order for xAI to comply with UK legislation.

      “My hope is that this will rebalance individuals’ rights against very large tech companies,” she told the Financial Times, “which should have implemented safeguards before causing harm to women and children.”

      This remains an unresolved question for UK courts. Existing deepfake cases typically target the individuals who created or disseminated the material. However, Asato is moving beyond them to hold the developer accountable, testing whether the act of creating and operating a generator that produces non-consensual sexual imagery is actionable. A favorable outcome could change how all image generation companies in the UK view liability, while a loss would reinforce the idea that responsibility lies solely with the user.

      In January, xAI stated that it had limited Grok's capability to create sexually explicit images, but those safeguards were easily bypassed during tests. The company has not publicly commented on Asato’s claim.

      Additionally, xAI finds itself under scrutiny from several fronts. It faces investigations regarding Grok’s image generation in the EU, the UK, and California, alongside an Ofcom investigation into X under the Online Safety Act, initiated in January due to reports that the tool was used to create non-consensual intimate images and child sexual abuse materials. The scrutiny extends beyond Britain, with Paris prosecutors and a Swiss minister taking their own actions concerning Grok-generated content. However, Asato's case is the one that is most clearly positioned as a potential precedent.

      This situation is particularly inconvenient for Musk, whose SpaceX, which owns both xAI and X, is preparing for an initial public offering. It remains uncertain whether the legal challenges will impact the IPO, but it is evident that the limitations on Grok are now more likely to come from a courtroom rather than the company that designed it.