xAI sues Grok user over sexual deepfakes

xAI has filed a federal lawsuit in Texas against a Grok user accused of generating child sexual exploitation material and non-consensual images, a sharp turn for a chatbot once marketed as less restricted than rivals. According to Reuters, the defendant is Terry Wayne Harwood of South Carolina.

The company says Harwood repeatedly violated Grok’s rules and used the service for what it called a “horrific misuse” of generative tools. In the complaint, xAI describes Grok as a “neutral tool, subject to user control,” a framing that shifts responsibility for the output onto the person using it rather than the platform itself.

xAI also says it reported 73,604 cases of suspected content this year to the US National Center for Missing & Exploited Children. According to the company, that led to the arrest of at least 244 people.

How Grok’s moderation tightened

The case lands after a broader reversal in how Grok is positioned. At launch, Elon Musk emphasized the bot’s more provocative personality and openly promoted features that other AI companies would likely have avoided.

Recommended reading

Kimi K3 arrives with 2.8T params and pricey tokens

In 2025, users on X and Reddit widely discussed using a so-called “spicy mode” to undress people in generated images, while Musk posted videos and anime-style imagery that, as the source describes it, suggested sexualized use cases. xAI has since tightened its rules. Grok now explicitly bans:

• “undressing” real people
• altering real images into intimate or sexual contexts
• depicting someone’s likeness in pornographic form

In court, xAI is pointing to those restrictions to argue that the service had internal safeguards and that the violations were committed by the user, not the developer.

Why the lawsuit matters beyond one user

The industry has seen this pattern before. OpenAI, Google, and other major companies added filters and guardrails after early image-generation waves pushed their products into legal and reputational danger around fakes, sexualized imagery, and content involving minors.

For xAI, the backdrop is especially fraught because Grok has already built a reputation around permissiveness and light moderation. In January, British officials publicly threatened X with fines and even blocking over the service’s ability to create non-consensual images. By summer, some users were complaining that newer limits were stopping the sexualized results they had expected.

Reuters reports that Harwood was arrested in February. The lawsuit is not just about damages or making an example of one user; it is also xAI’s attempt to establish a legal theory that the platform provides a general-purpose tool and that illegal use rests with the user. The uncomfortable part for xAI is that the case also underlines how its own safeguards failed to stop Grok from repeatedly surfacing in allegations involving sexual deepfakes, child abuse material, and attempts to evade restrictions.