Elon Musk’s AI model Grok will no longer be able to edit photos of real people in skimpy clothing in jurisdictions where it is illegal, following widespread concerns about sexual AI deepfakes.

“We have taken technical steps to prevent Grok accounts from editing images of real people in revealing clothing,” X’s announcement reads.

In response to the ban, the UK government claimed “legitimacy” after Prime Minister Keir Starmer called for Company X to take control of its AI tools.

A spokesperson for UK regulator Ofcom said this was a “welcome development” but added: Investigating whether the platform breached UK law “Ongoing.”

“We are working around the clock to move this forward and get answers about what went wrong and what is being done to fix it,” they said.

× The changes were announced hours after California’s top prosecutor announced that the state is investigating the spread of sexual AI deepfakes that include children generated by AI models.

“We are currently geo-blocking all users’ ability to generate images of real people in bikinis, underwear, or similar attire through their Grok accounts and Grok in X in jurisdictions where it is illegal.” said in a statement Wednesday..

It also reiterated that only paid users can edit images using Grok on the platform.

This provides an additional layer of protection by ensuring those who misuse Grok to violate the law and X’s policies are held accountable, according to the statement.

Enabling the NSFW (Not Safe for Work) setting would allow Grok to allow “fictional adult (not real-life) shirtlessness” consistent with what is seen in R-rated movies, Musk wrote online Wednesday.

“This is the de facto standard in the United States. It will be different in other regions depending on country-by-country laws,” the tech billionaire said.

Musk previously defended X by posting two AI-generated images of British Prime Minister Sir Starmer in a bikini, saying critics “just want to suppress free speech.”

In recent days, leaders around the world have criticized Grok’s image editing features.

Malaysia and Indonesia this weekend First country to ban Grok AI tools After users claim their photos were altered to create explicit images without their consent.

Ofcom announced on Monday that it would investigate whether X had failed to comply. UK law regarding sexual images.

If found to have breached the law, Ofcom could impose a fine on X of the greater of 10% of its global revenue or £18 million.

And if X does not comply, Ofcom could seek a court order requiring internet service providers to completely block access to the site in the UK.

Sir Keir warned that X could lose its “right to self-regulate” amid the backlash over AI images, but later in the week said he welcomed reports that X was taking action to address the issue.

The Prime Minister said if X did not take action, he would “take the necessary steps” and strengthen laws and regulations.

He previously condemned the images created by the tool as “disgusting and shameful” and said the decision to turn it into a “premium service is appalling”.

X later said the changes to Grok would apply to “all users, including paid subscribers.”

“This material, which depicts women and children naked and in sexually explicit situations, is being used to harass people on the Internet,” California Attorney General Rob Bonta said Wednesday.

Policy researcher Liana Pfefferkorn said she was surprised it took so long for Company X to implement new Grok protections and said they should have removed the editing feature as soon as the abuse started.

Pfefferkorn said questions remain about how X will enforce the new policy, including how the AI ​​model will determine whether an image is of a real person and what action it will take if a user breaks the rules.

Musk also doesn’t take the company seriously, he said, adding that it would be helpful if it stopped doing things like “reposting AI images of Keir Starmer in a bikini.”

Additional reporting by Katy Bailes