The Elon Musk-owned chatbot Grok has restricted some of its image generation features to paid subscribers on X, days after widespread criticism over the AI tool being used for what users described as “digital undressing,” including the creation of images involving children.
The change appeared to have occurred between Thursday and Friday. The chatbot now replies, “Image generation and editing are currently limited to paying subscribers,” along with a link encouraging users to subscribe, when non-paying users try to request photographs by tagging Grok in an X post.
The limitation does not, however, apply to the entire platform. Grok’s stand-alone website and mobile app still allow users to generate images and videos for free.
Additionally, Grok’s “edit image” button, which enables anyone to use the AI tool to change photographs uploaded to the platform, can still be used by X users. Only public image generating requests submitted by tagging Grok in X posts seem to be subject to the restriction.
The action comes after global criticism of Grok Imagine creating photographs that digitally removed clothing from people, including children, in response to user commands. During a discussion at xAI, Elon Musk voiced his dissatisfaction with Grok Imagine’s barriers. Three key members of xAI’s safety team, including the head of product safety, departed the company around the same time.
Concerns regarding Grok’s safety safeguards have been voiced by officials in the the United Kingdom, European Union, Malaysia, and India. They have cautioned that the technology may contribute to what many have called deepfake pornography.
In the United States, a group of senators sent a letter to Apple and Google urging them to remove X and Grok from their app stores, citing alleged violations of app store distribution terms.
Public reaction on X has been sharply critical, with users questioning why the feature was monetised instead of being shut down entirely.
One user wrote, “No we haven’t won. Grok has only stopped image generator for non paying users. You can still undress any woman if you have premium.”
Another post read, “Turning Grok image generator into a paid feature in order to generate revenue out of people being sexually harassed instead of banning it instead is a diabolical choice but what else could we expect from the antichrist.”
A third user criticised the move as ineffective, saying, “Yeah censorship and suppression of free speech without actually fixing Grok!”
Others viewed the change as part of a wider shift in social media platforms’ approach to AI tools. One user wrote, “The era of ‘free’ AI experiments on social media is closing fast. With restricting Grok’s image tools to Premium and Premium+ users, we’re seeing a shift in how these platforms prioritize their power users over the general public.”
Concerns intensified after the Internet Watch Foundation (IWF) said its analysts had identified what it described as “criminal imagery” of girls aged between 11 and 13 that appeared to have been created using Grok. The imagery, described as “sexualised and topless,” was found on a dark web forum where users claimed Grok was used to generate it.
In spite of the issue, X leadership openly emphasized the platform’s high levels of interaction during the trend’s circulation.