X says Grok will block bikini-style or revealing clothing image edits of real people
X has announced new limits on its Grok chatbot after public criticism and global scrutiny over how the tool handled image editing features.
X has announced new limits on the image-editing features of its chatbot Grok after weeks of public criticism over how the tool handled sensitive visual content. The company said it has changed its systems to stop Grok from editing images of real people into revealing clothing, including bikinis. X shared the update through its @Safety account, saying it has added technical controls to address misuse of the tool.
The company stated that the new restrictions apply to all users, including those who pay for access to Grok. According to X, no user will be allowed to use the chatbot to alter images of real individuals in ways that involve bikinis, underwear, or similar clothing. The platform said it took this step to prevent misuse that could lead to harm or violations of local laws.
Also read: Google Pixel 10a may launch sooner than expected: Check expected features, price and more
Changes to Image Access and Regional Limits
Along with the editing restrictions, xAI said it will move all image-generation tools behind a subscription paywall. This means users who do not pay for Grok will no longer be able to create images using the chatbot. X also said it will block the ability to generate images of real people in revealing clothing in regions where such content breaks local rules. These limits will apply both through the Grok account and through Grok within X.
The announcement came shortly after the state of California confirmed it had initiated an investigation into xAI and Grok. Authorities are reviewing how the company handled reports of AI-generated nudity and content involving minors. California Attorney General Rob Bonta referred to a report that reviewed about 20,000 images generated by xAI over a short period. The report claimed that more than half of those images showed people wearing minimal clothing, and some images appeared to involve children.
Also read: Apple to fine-tune Gemini and use it quietly behind Siri without Google branding: Report
What’s the Company’s Stance
In its statement, X said it follows a policy of zero tolerance toward child exploitation. The company said it bans high-priority content that violates its rules, including child sexual abuse material and non-consensual nudity. Earlier, Elon Musk said he was not aware of any cases in which Grok generated nude images of underage individuals. He later clarified that when the not-safe-for-work setting is turned on, Grok may allow limited nudity involving imaginary adult figures, not real people. He also said these rules depend on local laws.
Also read: 5 Spotify features that other music apps can’t match
Several countries have already taken action against Grok. Malaysia and Indonesia recently blocked access to the chatbot, citing safety concerns and the handling of sexual content created by AI. In the United Kingdom, media regulator Ofcom is investigating xAI and Grok. Officials there have also said they would support a similar ban if needed. X said it will continue to review its tools and policies as regulators examine the use of AI-generated content.