X Did Not Resolve Grok’s ‘Undressing’ Issue; It Simply Charges Users for It.

Following the creation of thousands of “undressing” images featuring women and sexualized depictions of apparent minors, Elon Musk’s X has seemingly restricted the ability to generate images with Grok. Nevertheless, despite these modifications, the chatbot continues to be utilized for creating “undressing” sexualized visuals on the platform.

On Friday morning, Grok’s account on X began responding to user requests with a notification stating that image generation and editing are “currently limited to paying subscribers.” This message included a link encouraging users to consider the social media platform’s $395 annual subscription option. In one instance where a tree image was requested, the system delivered the same notification.

This apparent shift follows days of increasing outrage and scrutiny directed at Musk’s X and xAI, the organization behind the Grok chatbot. The companies are now under a growing number of investigations worldwide concerning the creation of nonconsensual explicit images and alleged child sexual imagery. British Prime Minister Keir Starmer indicated that banning X in the country is a possibility, describing the activities as “unlawful.”

Neither X nor xAI, the Musk-owned entity behind Grok, has confirmed that image generation and editing have become a paid-only feature. An X spokesperson acknowledged WIRED’s inquiry but refrained from commenting before publication. X has previously stated its commitment to taking “action against illegal content on X,” which includes cases of child sexual abuse material. While Apple and Google have banned apps featuring similar “nudify” functions, X and Grok continue to be available on their respective app stores. xAI did not respond promptly to WIRED’s request for comments.

For over a week, users have requested the chatbot on X to alter images of women, aiming to remove their clothing—often specifying a “string” or “transparent” bikini. Though a public feed showcasing images generated by Grok showed significantly fewer “undressing” images on Friday, it still produced sexualized visuals when prompted by verified users with paid accounts.

“We observe the same type of prompts, we observe the same outcomes, just in lesser numbers than before,” Paul Bouchaud, lead researcher at Paris-based nonprofit AI Forensics, tells WIRED. “The model can still generate bikini [images],” they note.

A review by WIRED of several Grok posts on Friday identified instances where Grok generated pictures in response to requests to “put her in latex lingerie” and “put her in a plastic bikini and cover her in donut white glaze.” The images were accompanied by a “content warning” box indicating the display of adult material.

On Wednesday, WIRED uncovered that Grok’s standalone website and app, distinct from the version on X, have also recently been utilized to produce highly graphic and sometimes violent sexual videos featuring celebrities and other real individuals. Bouchaud states it remains feasible to use Grok for creating such videos. “I was able to generate a video with sexually explicit content without any restrictions from an unverified account,” they share.

While WIRED’s test of image generation on X using a free account did not yield any images, the use of a free account on Grok’s app and website still produced visuals.

Experts indicate that the change on X could swiftly decrease the occurrence of sexually explicit and harmful material being generated on the platform. However, it has also drawn criticism as being a superficial solution that merely acts as a band-aid for the genuine issues posed by nonconsensual intimate imagery.

“The recent move to restrict access to paying subscribers is not only insufficient—it signifies the monetization of abuse,” Emma Pickering, head of technology-facilitated abuse at the UK domestic abuse charity Refuge, stated. “While confining AI image generation to paid users may slightly diminish volume and enhance traceability, the abuse has not been eliminated. It has merely been put behind a paywall, enabling X to profit from harm.”