Elon Musk's artificial intelligence model Grok will no longer be able to edit photos of real people to show them in revealing clothing in jurisdictions where it is illegal, following widespread concerns about AI-generated sexualized deepfakes, UNN reports with reference to the BBC.
Details
"We have implemented technological measures to prevent the editing of images of real people in revealing clothing, such as bikinis, in the Grok account. This restriction applies to all users, including paid subscribers," reads an announcement on X, which operates the Grok AI tool.
The change was announced hours after California's attorney general said the state was investigating the spread of sexualized AI deepfakes, particularly of children, generated by an AI model.
"We are now geoblocking the ability for all users to create images of real people in bikinis, underwear, and similar clothing through the Grok account and on Grok on X in those jurisdictions where it is illegal," X said in a statement on Wednesday.
It also reiterated that only paid users would be able to edit images using Grok on the platform.
This will add an additional layer of protection, helping to ensure that those who attempt to abuse Grok to violate the law or X's policy will be held accountable, according to the statement.
With NSFW (not safe for work) settings enabled, Grok should allow "topless nudity of imaginary adults (not real ones)," which is consistent with what can be seen in R-rated movies, Musk wrote on the network on Wednesday.
"This is the de facto standard in America. This will vary in other regions, depending on the laws of each country," the tech multi-billionaire said.
Earlier, Musk defended X, posting messages that critics "just want to suppress free speech," as well as two AI-generated images of British Prime Minister Keir Starmer in a bikini.
Questions remain about how X will enforce its new rules, such as how the AI model will know if an image is of a real person, and what action it will take when users violate the rules, experts say.
Addition
In recent days, world leaders have criticized Grok's image editing feature.
Last week, Malaysia and Indonesia became the first countries to ban the AI tool Grok after users claimed that photos were altered to create explicit images without their consent.
British media regulator Ofcom said on Monday that it would investigate whether X had violated UK law regarding sexual images. Starmer warned that X could lose its "right to self-regulate" due to the backlash over AI-generated images, but later that week said he welcomed reports that X was taking steps to address the issue.