Following major global backlash across continents for publicly generating sexualized deepfakes targeting women including minors, Elon Musk’s AI chatbot Grok is now preventing non-paying users from generating or editing images.
Post the update, Grok on Friday responded to image altering requests with the message: “Image generation and editing are currently limited to paying subscribers. You can subscribe to unlock these features.”
As per a report by Reuters, Grok is still granting image requests but only from X users with blue checkmarks given to premium subscribers who pay $8 a month for features including higher usage limits for the chatbot.
The chatbot, which is accessed through Musk’s social media platform X, has in the past few weeks been granting a wave of what researchers say are malicious user requests to modify images, including putting women in bikinis or in sexually explicit positions.
While subscriber numbers for Grok aren’t publicly available, researchers examining the issue said there was a noticeable decline on Friday in the number of explicit deepfakes that Grok is now generating compared with just days earlier.
Countries protest, saying ‘subscription’ not the issue
The change which was announced by the platform after numerous countries like India, UK, France, Germany and Brazil threatened action against XAI for publicly generating sexualized deepfakes did little to satisfy regulators across all these countries.
“This doesn’t change our fundamental issue. Paid subscription or non-paid subscription, we don’t want to see such images. It’s as simple as that,” said Thomas Regnier, a spokesman for the European Union’s executive Commission. The Commission had earlier slammed Grok for “illegal” and “appalling” behavior.
India’s Ministry of Electronics and Information Technology (MeitY) had earlier this month sent X a formal notice directing the platform to take down obscene, sexualised images generated by Grok within 72 hours and also asked the tech platform to submit a detailed compliance report.
The legal team at X responded to the Indian ministry within the given window stating they acknowledge the seriousness of the issue. In their response, X further promised to act against users prompting Grok to generate non-consensual or obscene content, including permanently disabling accounts.
On January 9, the ministry sent the legal team at X a follow up notice, seeking specific details on how Grok is being trained, moderated and restricted to stop the generation of illegal content before it appears. MeitY asked X to outline concrete technical and procedural measures, rather than relying only on user penalties and content takedowns.
Britain echoed similar concerns. “It is insulting to the victims of misogyny and sexual violence,” a spokesperson for Prime Minister Keir Starmer said, noting that it shows that X “can move swiftly when it wants to do so. We expect rapid action,” he said, adding that “all options are on the table.”
Musk’s statement
Following the update on Friday, Elon Musk also posted on X to warn users saying that anyone who asks the AI to generate illegal content would “suffer the same consequences” as if they uploaded it themselves.
While XAI’s own acceptable use policy prohibits “depicting likenesses of persons in a pornographic manner”, but people have been using Grok to digitally undress people without their consent, Images of Catherine, Princess of Wales, were among many to have been digitally de-clothed by Grok users on X.
‘Edgier’ AI pitch adds to pressure on Musk
Grok was launched in 2023 as an alternative to more tightly moderated AI systems, with Musk positioning it as an “edgier” chatbot willing to push boundaries. Building on this narrative, last summer, xAI introduced an image-generation feature called Grok Imagine, which included a so-called “spicy mode” capable of producing adult content.
Critics argue that this positioning, combined with the public visibility of Grok’s outputs on X, amplified the harm caused by sexualised deepfakes. Unlike private image-generation tools, Grok’s content could be easily reshared, exposing victims to widespread online abuse.
(With inputs from Agencies)
