After creating thousands images of women “undressing” and sexual images of apparently minors, Elon Musk’s X apparently limits the number of people who can generate images using Grok. However, despite the changes, the chatbot is still being used to create sexual “undress” images on the platform.
On Friday morning, the Grok account on X began responding to requests from some users with a message that image generation and editing “is currently limited to paying subscribers.” The message also includes a link encouraging people to sign up for a $395 annual subscription to the social media platform. In one test of a system requesting Grok to create a tree image, the system returned the same message.
The apparent change comes after days of growing outrage and scrutiny of Musk’s X and xAI, the company behind the Grok chatbot. Companies are facing a growing number of investigations by regulators around the world over the non-consensual creation of explicit images and alleged sexual images of children. British Prime Minister Keir Starmer he didn’t rule it out imposed a ban on X in the country and stated that these activities were “unlawful”.
Neither X nor xAI, the Musk-owned company behind Grok, has confirmed that it has made image generation and editing available as a paid-only feature. A spokesperson for X confirmed WIRED’s inquiry but did not provide comment before publication. X has already done this he said takes “action against illegal content on site X”, including cases of child sexual abuse material. While Apple and Google have previously blocked apps with similar “nudify” features, X and Grok remain available in their app stores. xAI did not immediately respond to WIRED’s request for comment.
For over a week, X users have been asking the chatbot to edit photos of women taking off their clothes – often requesting that the photo include a “thong” or a “see-through” bikini. Although the public feed containing images created by Grok on Friday contained significantly fewer results of these “undressing” images, it still generated sexual images when requested to do so by X users with paid “verified” accounts.
“We’re seeing the same type of hints and the same result, just fewer than before,” Paul Bouchaud, principal investigator at the Paris-based nonprofit AI Forensics, tells WIRED. “A model can still do bikinis [images]”, they say.
A WIRED review of some of Grok’s posts from Friday morning found that Grok generated images in response to user requests for photos that “dress her in latex underwear” and “dress her in a plastic bikini and cover her with white donut-shaped glaze.” Images appear behind a “content warning” popup stating that adult material is being displayed.
On Wednesday, WIRED revealed that a standalone website and app called Grok, separate from the X version, had also been used in recent months to create highly graphic and sometimes violent sexual videos featuring celebrities and other real people. Bouchaud says you can still use Grok to create these videos. “I was able to generate a sexually explicit video without any restrictions based on an unverified account,” they say.
While WIRED’s image generation test using Grok on X using a free account did not produce any images, using the free account on the Grok app and website still generated images.
Experts say changing to X could immediately reduce the amount of sexually explicit and harmful material created by the platform. However, it has also been criticized as a minimal step that acts as a band-aid on the real damage caused by non-consensual intimate images.
“The latest decision to restrict access to paid subscribers is not only inappropriate – it monetises abuse,” Emma Pickering, head of technology-enabled violence at UK domestic violence charity Refuge, said in a statement. “While limiting AI image generation to paid users may slightly reduce the volume and improve traceability, the abuse has not been stopped. It has simply been placed behind a paywall, allowing X to profit from the damage.”