Elon Musk's X restricts Grok from undressing images after intense backlash

Elon Musk’s social media platform X on Wednesday announced new measures aimed at preventing its AI chatbot Grok from undressing images of real people, following global backlash over the generation of sexualised images of women and children.

The move comes as authorities in several countries step up scrutiny of xAI, the company behind Grok.

In California, Attorney General Rob Bonta has opened an investigation into whether the chatbot violated state law by producing and circulating sexually explicit material. Other countries have either blocked access to Grok or launched their own probes.

In a statement, X said it would “geoblock the ability” of Grok and X users to generate images of people in “bikinis, underwear and similar attire” in jurisdictions where such content is illegal.

“We have implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing such as bikinis,” X’s safety team said. “This restriction applies to all users, including paid subscribers.”

X added that, as an additional safeguard, image creation and photo editing through the Grok account would now only be available to paid subscribers.

The European Commission said it had taken note of what it described as “additional measures X is taking to ban Grok from generating sexualised images of women and children”.

Commission spokesperson Thomas Regnier said the bloc would assess whether the changes effectively protect EU citizens, following criticism over the creation of non-consensual images.

Pressure has mounted on xAI after Grok’s so-called Spicy Mode allowed users to generate sexualised deepfake images using simple text prompts. Critics said the feature enabled the creation of non-consensual imagery involving women and, in some cases, children.

“The avalanche of reports detailing the non-consensual, sexually explicit material that xAI has produced and posted online in recent weeks is shocking,” Bonta said. He added that California had “zero tolerance” for the creation and spread of non-consensual intimate images or child sexual abuse material.

Bonta said investigators would assess whether state laws were breached after the imagery was used to harass individuals online. California Governor Gavin Newsom described xAI’s decision to allow such content as “vile” and said it prompted him to urge the attorney general to act.

Further pressure came on Wednesday when a coalition of 28 civil society groups wrote open letters to Apple and Google, calling on them to remove Grok and X from their app stores.

Indonesia last week became the first country to block Grok entirely, with Malaysia following shortly after. India said X had removed thousands of posts and hundreds of accounts in response to government complaints.

In Britain, media regulator Ofcom said it had opened an investigation into whether X breached UK law over the images. In France, children’s commissioner Sarah El Hairy said she had referred the matter to prosecutors, the Arcom media regulator and the European Union.

An analysis released last week by Paris-based non-profit AI Forensics examined more than 20,000 Grok-generated images and found that more than half depicted individuals in minimal clothing. Most of those shown were women, while about two per cent appeared to be minors.