Exclusive-SpaceX warns that inquiries into sexually abusive AI imagery may hurt market access

By Raphael Satter, Echo Wang and Jeffrey Dastin

MEMPHIS, Tennessee, April 23 (Reuters) – The multiple investigations into xAI’s creation and dissemination of sexually abusive imagery may lead the company to lose access to certain markets, parent company SpaceX warned in a prospectus reviewed by Reuters.

In a section on risk factors, the S-1 regulatory filing said a number of agencies around the world were “actively investigating and making inquiries relating to social media or the use of AI” in relation to advertising, consumer protection and the distribution of harmful content, among other matters.

The news comes after SpaceX hosted analysts at its Colossus supercomputer in Memphis, Tennessee, on Thursday, gearing up for its $1.75 trillion IPO expected this summer. U.S. securities law requires companies to disclose such risk factors, alerting investors to potential pitfalls while also helping protect companies against future legal liability. The disclosures do not necessarily mean each listed outcome is expected to occur.

One challenge SpaceX highlighted was that it faced “allegations that our AI products were used to create nonconsensual explicit images or content representing children in sexualized contexts,” the S-1 document said. Such regulatory inquiries could expose SpaceX to lawsuits, liability and government action – “including loss of access to certain markets, which has occurred in the past,” the document stated.

SpaceX and xAI did not immediately respond to requests for comment. It was not clear whether potential regulatory action could prevent SpaceX as a whole from accessing certain markets or just its subsidiary, xAI, specifically. 

WORLDWIDE SCRUTINY OVER GROK IMAGES

Though the regulatory filing’s risk factors gave as an example a probe launched by the Irish Data Protection Commission in February, xAI has faced scrutiny worldwide over an explosion of sexualized images. The content, which was particularly visible in late 2025 and early 2026, featured images of nearly naked women and children on X, the company’s social media platform.

XAI said in January that it had added measures to block user requests for sexualized images of real people, and it said it stops users from generating such content in jurisdictions where that is illegal.

The images – which were generated by xAI’s in-house chatbot, Grok – had shown women and sometimes minors in revealing bikinis or underwear, or edited into degrading or gruesome poses. The pictures caused widespread alarm around the world; one group of researchers estimated there were about 3 million sexualized images, while U.S. lawmakers demanded that Google owner Alphabet and Apple yank Grok and X from their app stores. SpaceX CEO Elon Musk said around that time that he knew of “literally zero” naked underage images made by Grok.

A variety of investigations – launched earlier in Canada, Britain, Brazil, California and elsewhere – are ongoing. In France, for example, Musk on Monday flouted a legal summons to answer questions from prosecutors concerning allegations of algorithmic abuse, fraudulent data extraction and complicity in the distribution of child sexual abuse material. 

STAKES HIGH AS PROBES CONTINUE

The S-1 warning around market access illustrates the stakes of the various investigations into xAI, especially ones around the AI generation of alleged child sexual abuse images and nonconsensual sexual images of women. Creation of such imagery can be a crime in some jurisdictions, and its dissemination is an emotive issue that can rapidly mobilize public opinion.

XAI’s curbs on Grok appear to have slowed but not stopped the flow of abusive material. In February, Reuters reported that Grok was generating sexualized imagery of people even when users explicitly warned the chatbot that the subjects of those images did not consent. Last week, NBC News found that Grok was still publicly generating sexualized images, including of actors and pop stars.

X has been banned before in various jurisdictions, including in 2024 in Brazil, where the site was blocked following its refusal to comply with a judge’s order. The company later relented and the ban was lifted.

(Reporting by Raphael Satter in Memphis, Tennessee, Echo Wang in New York and Jeffrey Dastin in San Francisco; Writing by Raphael Satter; Editing by Kenneth Li and Matthew Lewis)