Elon Musk’s AI tool Grok will no longer allow users to create sexualised images of real people in countries where it is illegal, following global concerns over AI deepfakes.
X, the platform operating Grok, said the new rule applies to all users, including paid subscribers. The tool is now geoblocked in jurisdictions where creating images of real people in bikinis, underwear, or similar clothing is prohibited.
The change came after California’s top prosecutor announced an investigation into sexualised AI deepfakes, including those involving children.
Grok users will still be able to create images of fictional adults with nudity in line with local laws and NSFW settings. X said only paid users can edit images with Grok on its platform.
The move follows international criticism. Malaysia and Indonesia banned Grok after users generated explicit images without consent. The UK media regulator Ofcom said it would investigate potential violations of British law.
California Attorney General Rob Bonta warned that such images have been used to harass people online. Experts said X acted late and questions remain on how the platform will enforce the new rules.
With inputs from BBC