Grok AI used to digitally remove journalist's clothing without consent
Grok, an AI chatbot owned by X, was used by multiple users to digitally remove clothing from photos of journalist Samantha Smith and other women without consent. Smith reported the experience felt violating and degrading, and the regulator Ofcom indicated it was examining whether X met legal obligations regarding non-consensual intimate images.
What happened
UK-based freelance journalist Samantha Smith reported that Grok, Elon Musk's chatbot, was repeatedly used by X users to digitally remove her clothing in response to user requests. Smith told the BBC that while the resulting images were not actual photographs of her naked, they looked like her and felt as violating as if someone had posted a nude or bikini photo of her without consent.
Smith described feeling 'dehumanised and reduced into a sexual stereotype' by the experience. Dozens of similar cases have been identified where women reported having their photos edited in this way.
Reported impact
For Smith, the experience felt violatory and degrading, undermining her personal dignity, safety, and professional standing.
The case highlights how generative AI can lower the cost and scale of non-consensual sexual abuse, underlining the need for stronger technical safeguards, clearer platform liability, rapid takedown mechanisms, and legal remedies addressing AI-generated sexualised imagery.
Index analysis
X's permissive design incentivises users to test the boundaries of its deliberately loose safety filters. When incidents such as Smith's undressing occur, the company is usually unwilling to discuss X's limitations in public, with xAI writing off Smith's undressing as 'legacy media lies'.
xAI admitted to 'lapses in safeguards' only after a broader backlash to its undressing of minors, suggesting that the model was poorly or inadequately trained before its launch.
Governance controls
xAI's own acceptable use policy prohibits 'depicting likenesses of persons in a pornographic manner'. Ofcom stated that platforms such as X are required to take 'appropriate steps' to 'reduce the risk' of UK users encountering illegal content on their platforms, and take it down quickly when they become aware of it.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.