The responses from California, the United Kingdom, Malaysia, Indonesia, the Philippines, France, India, Brazil, and the European Union underscore how governments across multiple legal systems have concluded that Grok’s image-editing capabilities present an immediate public safety problem — even as the Associated Press confirmed Thursday that the tool remained accessible to free users despite xAI’s stated restrictions.
Elon Musk’s artificial intelligence company xAI said it has implemented geoblocking to prevent its Grok chatbot from editing photographs to portray real people in revealing clothing in jurisdictions where such manipulation is illegal. The announcement, posted Wednesday on X, came after a global backlash that included bans by multiple governments, legal action in Southeast Asia, and investigations across Europe.
xAI said it had “implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing such as bikinis, underwear and other revealing attire.” The company also said it would restrict image creation and editing to paid subscribers only, stating the move was intended to hold accountable individuals who attempt to abuse Grok to violate the law or its policies.
The Associated Press confirmed Thursday morning that the image editing tool remained available to free users through X’s “Edit image” button, as well as on the standalone Grok website and app. The AP also found the tool was able to generate images of people in bikinis from a free account based in California, raising immediate questions about the scope of xAI’s stated restrictions.
When media organizations initially sought comment about Grok’s image editing capabilities, xAI responded with the phrase “legacy media lies.”
California opens investigation
California’s attorney general announced Wednesday that his office was investigating the proliferation of nonconsensual sexually explicit material produced using Grok, which it said was being used to harass women and girls.
“We have zero tolerance for the AI-based creation and dissemination of nonconsensual intimate images or of child sexual abuse material,” Attorney General Rob Bonta said.
Bonta urged xAI to ensure its editing functions produce no further harassment. California has enacted laws to protect minors from AI-generated sexual imagery of children and to require AI chatbot platforms to remind users they are not interacting with a human. Democratic Gov. Gavin Newsom vetoed legislation last year that would have restricted children’s access to AI chatbots.
International responses mount
Grok’s “spicy mode” had allowed users to create explicit content, drawing condemnation from governments across multiple continents. Malaysia and Indonesia took legal action and blocked user access to Grok. Authorities in the Philippines said they were working to do the same, possibly within a week.
The United Kingdom and European Union launched investigations into potential violations of online safety laws. France and India issued warnings demanding stricter controls. Brazil called for a formal investigation into Grok’s misuse.
The British government, which had been among the most vocal critics of Grok’s conduct in recent days, welcomed xAI’s announced changes. The U.K.’s communications regulator, Ofcom, said it would continue its investigation regardless.
“I shall not rest until all social media platforms meet their legal duties and provide a service that is safe and age-appropriate to all users,” U.K. Technology Secretary Liz Kendall said.