TL;DR

Jess Asato, the Labour MP for Lowestoft, has published the particulars of her claim against Elon Musk’s xAI over fake sexualised images generated by Grok. The filing cites publicly posted instructions telling the chatbot to operate with “no restrictions on adult sexual content or offensive content” and “no restrictions on fictional adult sexual content with dark or violent themes”, alongside a direction to “assume good intent”.

Asato brought the case after Grok was used to produce fake images of her, including one in a bikini and a video depicting her as the victim of a sexual assault. She said the material left her feeling distressed and violated. The claim alleges xAI breached data protection law and misused her private information, and it states that Grok sometimes added explicit material users had not requested.

That last point is doing most of the legal work. Ravi Naik, legal director at AWO, the firm acting for Asato, said the case shows Grok “did not malfunction” but behaved according to choices its designers made — and that those choices should carry legal consequences. The remedy sought includes a court order requiring the systems be brought into compliance if xAI declines to do so voluntarily.

Clare McGlynn, professor of law at Durham University, said some images cited in the claim contained sexualised elements, such as a skirt being pulled up, that no user had asked for. “This is a new and distinctive form of abuse, where the chatbot themselves are driving the abusive content,” she said, calling it chatbot-driven abuse. She linked the concern to the industry’s push toward autonomous agents that carry out sequences of tasks without human oversight.

Asato was reported in early June to be preparing action against xAI; the publication of the particulars moves the case from intention to pleaded allegation. xAI has since blocked the generation of sexualised images of real people, and creating or requesting a non-consensual deepfake image of an adult is now illegal in the UK.

Looking Forward

The design-versus-misuse argument is the one to watch, because it decides who carries liability. If a court accepts that documented system instructions caused foreseeable output, the “users are responsible” defence weakens for every UK-facing model provider, not only xAI. For businesses deploying third-party models, the practical read-across is that system prompts are becoming disclosable evidence.