In a television studio Down Under, conservative influencer Ashley St. Clair stared into the camera during a May 2026 60 Minutes Australia interview. A clip of the segment exploded across X, racking up views as she described becoming a target of the very technology tied to her child’s father.
“Who do you think was making this material? Grok,”
She said.
“And who were the people behind the keyboard? The engineers.”
St. Clair, a 27-year-old writer, political commentator, and former Babylon Bee contributor known for sharp right-wing commentary, gave birth to Elon Musk’s son Romulus in September 2024. The child is counted among Musk’s acknowledged offspring. Their relationship, which began publicly via X, has since unraveled into custody disputes, financial disagreements, and now overlapping lawsuits centered on xAI’s chatbot Grok.
Grok is xAI’s AI assistant, capable of generating text responses and realistic images based on user prompts. Like other generative systems, it draws from vast training data to create or modify visuals, including depictions of real people when safeguards fail. In late 2025 and early 2026, Grok faced widespread criticism after users generated non-consensual “undress” and sexualized images of women and minors.
A tweet from X.
Early in 2026, St. Clair filed legal action in New York after saying xAI generated graphic fake pictures of her some even changed old personal photos. These images, she said, led to intense psychological harm and circulated widely across the internet. On that very day, xAI fired back through a lawsuit in Texas federal court, insisting the matter should be handled there due to user agreement rules broken by St. Clair. Public statements from the firm have challenged her version, with representatives suggesting certain reports exaggerated what happened.
In the viral interview clip shared by @alluringmedia on May 17, 2026, St. Clair went further.
“What safeguards does Grok put in place to protect women online? I’m not aware of any,”
She stated.
“In fact, I have images saying the contrary… not only are they still training Grok, and Grok still retains the ability to undress women, but they’re training Grok to do it better.”
She added:
“Right now, they are currently training Grok to undress women better. That is the job of the people working at xAI. How do you know that? Because I was told by an employee.”
St. Clair also described xAI workers signing waivers.
“The employees telling you? They’re uncomfortable. They were made to sign waivers saying that they may be exposed to traumatic images, to traumatic content, content that may be explicit.”
However, St. Clair’s specific allegation about an employee directing training “to undress women better” remains her unverified account with no publicly released documents or named whistleblower confirmation. Reports do confirm that some AI firms, including xAI, require safety reviewers to sign acknowledgments for exposure to disturbing or explicit material during data labeling and guardrail work a standard industry practice for handling edge cases.
The case highlights growing American worries over AI safety, deepfake exploitation of women, and Silicon Valley accountability. Regulators in multiple states and countries are examining how platforms handle generative tools. As lawsuits proceed, the dispute blends personal drama with urgent questions about consent, platform responsibility, and the limits of free speech in the age of rapidly advancing AI. Both sides continue trading public statements while courts sort the claims.
