X says Grok will no longer edit images of real people into bikinis

X says it is changing its policies around Grok’s image-editing abilities following a multi-week outcry over the chatbot repeatedly being accused of generating sexualized images of children and nonconsensual nudity. In an update shared from the @Safety account on X, the company said it has “implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing such as bikinis.”

The new safeguards, according to X, will apply to all users regardless of whether they pay for Grok. xAI is also moving all of Grok’s image-generating features behind its subscriber paywall so that non-paying users will no longer be able to create images. And it will geoblock “the ability of all users to generate images of real people in bikinis, underwear, and similar attire via the Grok account and in Grok in X” in regions where it’s illegal.

The company’s statement comes hours after the state of California opened an investigation into xAI and Grok over its handling of AI-generated nudity and child exploitation material. A statement from California Attorney General Rob Bonta cited one analysis that found “more than half of the 20,000 images generated by xAI between Christmas and New Years depicted people in minimal clothing,” including some that appeared to be children.

In its update, X said that it has “zero tolerance” for child exploitation and that it removes “high-priority violative content, including Child Sexual Abuse Material (CSAM) and non-consensual nudity” from its platform. Earlier in the day, Elon Musk said he was “not aware of any naked underage images generated by Grok.” He later added that when its NSFW setting is enabled, “Grok is supposed [sic] allow upper body nudity of imaginary adult humans (not real ones) consistent with what can be seen in R-rated movies on Apple TV.” He added that “this will vary in other regions” based on local laws.

Malaysia and Indonesia both recently moved to block Grok citing safety concerns and its handling of sexually explicit AI-generated material. In the UK, where regulator Ofcom is also investigating xAI and Grok, officials have also said they would back a similar block of the chatbot.

Have a tip for Karissa? You can reach her by email, on X, Bluesky, Threads, or send a message to @karissabe.51 to chat confidentially on Signal.