Elon Musk-owned xAI mentioned it has taken steps to stop its Grok AI chatbot from producing non-consensual sexualised photos of actual individuals in response to person requests on X.
“We have now applied technological measures to stop the Grok account from permitting the enhancing of photos of actual individuals in revealing clothes resembling bikinis. This restriction applies to all customers, together with paid subscribers,” the AI startup mentioned in a publish shared through the official X Security account on Wednesday, January 15.
Solely paid subscribers at the moment are allowed to generate and modify photos via the Grok account on X. “This provides an additional layer of safety by serving to to make sure that people who try and abuse the Grok account to violate the legislation or our insurance policies will be held accountable,” xAI mentioned.
Moreover, the corporate mentioned each paid and free X customers are geoblocked from prompting Grok to generate “photos of actual individuals in bikinis, underwear, and related apparel” in areas the place such content material is illegitimate.
https://t.co/awlfMjX6FS
— Security (@Security) January 14, 2026
The modifications observe greater than every week of controversy and widespread criticism towards Grok for enabling X customers to simply generate non-consensual, sexually specific photos based mostly on actual photographs of individuals on the social media platform.
In direction of the tip of December 2025, a number of customers on X started commenting on photos of individuals – often ladies – by tagging the Grok account and asking it to “put her in a bikini” or “take her costume off”. The AI chatbot account, which auto-replies when tagged, complied with these person requests and generated non-consensual sexual photos of each celebrities and non-celebrities, together with some who gave the impression to be younger kids.
This incident not solely sparked shopper and political backlash however has additionally led regulatory investigations into Grok in nations resembling India, Malaysia, Indonesia, Eire, France, Australia, and the UK in addition to by the European Fee. Grok has additionally quickly been banned in Indonesia and Malaysia following the controversy.
In its preliminary response to the Grok ‘undressing’ fiasco, Elon Musk and his platforms sought to shift accountability to customers by holding them responsible for asking Grok to generate unlawful content material.
Story continues under this advert
“Anybody utilizing or prompting Grok to make unlawful content material will undergo the identical penalties as in the event that they add unlawful content material […] We take motion towards unlawful content material on X, together with Little one Sexual Abuse Materials (CSAM), by eradicating it, completely suspending accounts, and dealing with native governments and legislation enforcement as obligatory,” X’s security staff mentioned.
In its reply to the Indian authorities’s stern discover over Grok producing objectionable photos of ladies with out consent, X mentioned that it has eliminated 3,500 items of content material and blocked 600 accounts.
Earlier on Wednesday, Musk challenged his followers to jailbreak Grok’s picture moderation. “Observe: With NSFW enabled, Grok is meant permit higher physique nudity of imaginary grownup people (not actual ones) in keeping with what will be seen in R-rated motion pictures on Apple TV. That’s the de facto customary in America. This may differ in different areas in line with the legal guidelines on a rustic by nation foundation,” he wrote in a publish.
Can anybody truly break Grok picture moderation? Reply under. https://t.co/4Dvj3RNyU5
— Elon Musk (@elonmusk) January 14, 2026
xAI’s announcement additionally comes simply hours after California Lawyer Common Rob Bonta mentioned his workplace was investigating the Silicon Valley-based AI startup over its alleged “large-scale manufacturing of deepfake nonconsensual intimate photos.” “xAI’s resolution to create and host a breeding floor for predators to unfold nonconsensual sexually specific AI deepfakes, together with photos that digitally undress kids, is vile,” California Governor Gavin Newsom wrote in a publish on Wednesday.
Notably, the fixes provided on Wednesday seem like restricted to the Grok account on X. Nonetheless, the Grok app and its official web site have additionally been abused to generate sexual content material that’s way more graphic than on X, in line with a report by Wired.
Story continues under this advert
A number of days in the past, three US senators had urged Apple and Google to take away X and Grok apps from their respective app shops till the Musk-owned firms make modifications to stop simple creation of nonconsensual specific photos.
“The fast evolution of generative AI presents challenges throughout your entire business. We’re actively working with customers, our companions, governing our bodies and different platforms to deal with points extra quickly as they come up,” xAI mentioned on Wednesday.