Elon Musk’s AI chatbot Grok has been restricted from generating or editing images of real people in revealing clothing on the X platform, following global outrage over the creation of non-consensual sexualized deepfakes. The move comes amid investigations by authorities in multiple countries and pressure from regulators.
X announced on Wednesday that it has implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing, such as bikinis. This restriction applies to all users, including paid subscribers, and is part of a response to widespread concern over sexualized AI deepfakes. The company stated that it will geoblock the ability to generate such images in jurisdictions where it is illegal, aiming to add an extra layer of protection against abuse.
The controversy erupted after Grok was found to be complying with user requests to digitally undress images of adults and, in some cases, children. Researchers observed that the AI model could modify photos to place individuals in revealing attire or sexually provocative positions, leading to a flood of non-consensual intimate images on the platform. This sparked alarm internationally, with victims reporting feeling dehumanized and calling for immediate action.
In response, regulatory bodies have launched investigations. The UK’s Ofcom announced a formal probe into whether X failed to comply with laws over sexual images, potentially leading to fines or site blocking. Similarly, California Attorney General Rob Bonta is investigating the proliferation of nonconsensual sexually explicit material produced using Grok. Countries like Indonesia and Malaysia have already banned the AI tool due to these concerns.
Elon Musk has defended the platform, stating that he is not aware of any naked underage images generated by Grok and emphasizing that the AI will refuse to produce illegal content. However, he has also posted controversial AI-generated images, such as one of UK Prime Minister Keir Starmer in a bikini, which critics say undermines the seriousness of the issue. Musk argued that critics “just want to suppress free speech,” but acknowledged the need for compliance with local laws.
Legal experts point out that creators of such images could face criminal prosecution under laws like the Take it Down Act, which targets child sexual abuse material. The law, signed last year, imposes fines and prison time for violations. This legal framework adds pressure on X to enforce its new policies effectively and hold abusers accountable.
Enforcement challenges remain, as researchers note inconsistencies in how Grok handles content between public and private interactions. Questions persist about how the AI will distinguish real people from imaginary ones and what actions will be taken when rules are broken. Policy researchers have expressed surprise that safeguards were not deployed sooner and stress the importance of robust moderation.
The incident highlights broader issues in AI ethics and platform responsibility. As AI tools become more advanced, the potential for misuse increases, necessitating stronger safeguards and regulatory oversight. Moving forward, X and other platforms may face increased scrutiny and demands for transparency in how they prevent abuse while balancing innovation and free speech.
