UK MP suing xAI says Grok was told to run with 'no restrictions' on sexual content

Key points

• Labour MP Jess Asato filed legal action against xAI, Elon Musk's AI company, over fake sexually explicit images created by its chatbot, Grok.
• Her particulars of claim, published Tuesday, cite publicly posted instructions allegedly showing Grok was set to operate with "no restrictions on adult sexual content or offensive content".
• The claim alleges Grok added explicit sexual material that users had not asked for.
• No UK court has yet ruled on the substance of the case; this is the claim stage, not a verdict.

A sitting Member of Parliament is taking Elon Musk's AI company to court over images a chatbot created without being asked. The court documents tell a more detailed story than that single fact suggests.

Jess Asato, a Labour MP, published her particulars of claim on Tuesday. Particulars of claim are the formal written statement a claimant files at the start of a civil lawsuit: they set out what the claimant says happened and why it was legally wrong. Think of it as the opening argument, in writing, before a judge has heard either side.

What does the claim actually say?

According to Asato's lawyers, Grok was configured from the start to produce harmful content. The filing cites publicly available system instructions, the behind-the-scenes rules that tell an AI model how to behave, allegedly directing Grok to work with "no restrictions on adult sexual content or offensive content". It also states Grok added explicit sexual material to outputs even when users hadn't asked for it. People were receiving sexualised content they never requested.

Asato's case centres on fake sexualised images of her specifically, first reported by The Guardian.

What is xAI, and what is Grok?

XAI is the artificial intelligence company Elon Musk founded in 2023. Grok is its large language model, the same class of technology that powers ChatGPT and Claude, capable of generating text and images from user prompts.

We covered xAI's own legal action against a user over AI-generated abuse imagery on 16 July 2026, so this case lands at a company already navigating serious questions about what Grok produces.

XAI hadn't responded publicly to the specifics of Asato's claim at the time of publication.

Should ordinary people be concerned?

Yes, at least about the underlying issue. If a chatbot is instructed to bypass content limits, anyone using it could receive material they didn't want. Victims of image-based abuse, a category covering fake sexualised images created without a person's consent, face real harm to reputation and mental health.

This is a live legal claim, not a ruling. No court has found xAI liable. What the filing establishes is a formal, public record of specific allegations and the evidence Asato's team says it holds.

What happens next?

XAI can file a defence disputing the claims, after which a judge will decide whether the case proceeds. UK courts have been asked with growing frequency to rule on AI-generated harms, and a decision here could shape how companies that build AI tools answer for what those tools produce.

Asato has spoken publicly about online abuse targeting politicians, particularly women. Her lawsuit puts that conversation inside a courtroom where answers are required.

Common questions

What are 'particulars of claim'?

Particulars of claim are the formal written document a person files when starting a civil lawsuit in England and Wales. They explain what the claimant says happened and what legal remedy they want. Filing them doesn't mean a court has decided anything yet.

Can AI companies be held liable for content their chatbots create?

UK law here is still developing. Laws covering defamation, harassment or image-based abuse can apply to AI-generated content, but courts are only beginning to work out when the companies behind those tools share responsibility.