In mid-December, Mollie posted a photo on X that showed her squatting in front of a Christmas tree in leggings and a long-sleeved shirt. “Are we getting into the Christmas spirit?” the 24-year-old captioned the post. A few weeks later, Mollie saw one of her followers reply underneath. “Put her in a micro bikini from this angle,” the user wrote, tagging the X AI chatbot Grok. Grok complied, showing an AI-generated image of Mollie squatting in the same position and wearing a thong bikini instead of workout gear.
As a fairly prominent adult creator — she has nearly 80,000 followers on X alone — Mollie is used to having random followers edit her content without her consent. Last year, she noticed some of her fans digitally altering images to make her bald, which she thought made her look somewhat like Matt Damon. But someone using AI to remove her clothes felt different. Her shock over how real the manipulated photo looked quickly gave way to mortification, then horror. “It was scary and it was uncomfortable to have that power asserted over you,” she says. “I’ve been sexually assaulted in the past, and it almost felt like a digital version of that.”
She told Grok to take the photo down. “I understand it violated your consent,” Grok responded. “I won’t create images like that going forward.” The photo, however, is still up on X, and Mollie continues to see women she follows on the platform complain about having their own images altered in a sexually suggestive way without their consent. “It is unfathomable to me that people are allowed to do this to women,” she says.
For the past week, X has been flooded by complaints from women begging Grok to stop digitally altering their photos to make them appear nude or put them in sexually compromising positions. The anti-Grok backlash started when X users pointed out that Grok had generated AI images of two scantily clad girls who appeared to be underage. On December 28, the Grok account issued a statement apologizing for the “incident,” referring to it as “a failure in safeguards” that “violated ethical standards and potentially US laws on CSAM.” X’s Safety account also issued a statement, saying, “We take action against illegal content on X, including Child Sexual Abuse Material (CSAM), by removing it, permanently suspending accounts, and working with local governments and law enforcement as necessary.”
X head Elon Musk, however, has been relatively quiet on the topic. He did reply to a user likening Grok creating inappropriate images to blaming a pen rather than the person holding it for writing something bad. “Anyone using Grok to make illegal content will suffer the same consequences as if they upload illegal content,” Musk wrote. (Musk and X did not immediately respond to The Cut’s requests for comment.)
This is not the first time Grok or other AI chatbots have been used for nefarious purposes. Last year, the Verge published a report alleging that it had used Grok to create nude deep fakes of Taylor Swift. Sex workers and adult-content creators have also been subject to this type of treatment for years, with 4chan trolls orchestrating a 2024 campaign to use AI to deface and digitally alter their images. Deep fakes obviously threaten OnlyFans creators’ bottom line — but they’re also a means of shaming and controlling women, says Mollie.
“There’s such a huge difference between me saying, ‘Hey, if you pay X amount, then you can see these photos,’ versus someone having the option to do that via Grok just to assert their power and humiliate you publicly,” she says. “The fact that they don’t have consent is what they like about it. It’s just a way to make you uncomfortable.”
U.S. federal law now technically prohibits the publication of “intimate images, including ‘digital forgeries’ (i.e., deep fakes),” thanks to President Trump signing the Take It Down Act last May. Social-media companies are required to take down reported content within 48 hours, though they have a year to come up with a process to comply with those requests, and the definition of what constitutes sexually explicit content under the law may not encompass scantily clad rather than fully nude women. Many states also have laws on the books specifically prohibiting nonconsensual, AI-generated sexual content, and some AI models have specific frameworks in place to try to prevent users from exploiting them to create child sexual-abuse material (CSAM).
Click here to preview your posts with PRO themes ››
But Grok appears to have made it exceptionally easy to generate sexualized imagery of women and girls without their consent. One 27-year-old content creator, a woman who goes by Miau, says one of her followers edited a post-workout video of her dancing in a gym locker room. “When I saw the tags for Grok, I was confused. I thought maybe it was a meme,” she says. “I thought, Surely Grok would not make porn for these people.” She was disgusted when she realized the video had been altered to remove her clothes, and though she reported the accounts spreading the post, she says X didn’t take any action. “It gives you a very dreadful feeling, seeing yourself do something that you never actually did,” she says.
One of the most high-profile examples of women speaking out against the trend is Ashley St. Clair, a conservative author who alleges that Musk is the father of her toddler and has sued him for sole custody of the child. St. Clair says that she learned Grok had been altering her images without her consent when a friend sent her an AI-edited photo showing her in a bikini. “I was disgusted and horrified,” she tells me.
St. Clair replied to Grok publicly, requesting that it remove the photo. What followed was a deluge of X users posting even more digitally altered photographs of her in the replies to her post, including one showing her bent over and partially nude — “my toddler’s backpack was in the background,” she says angrily — and another that was taken when she was just 14 years old and manipulated to show her undressed.
“With each of these, I responded to Grok, and I said, ‘Please confirm that I don’t consent to this,’” St. Clair says. “And with each one, Grok said, ’I confirm you already made this request, and I won’t produce them.’ And it continued producing them.” St. Clair was particularly horrified by one Grok-altered image of a very young child wearing a Bluey dress and covered in a substance resembling semen. “I don’t know who this little girl is, but I’ve been thinking about her and her parents all day,” she says. (The images have since been removed from X, but St. Clair provided screengrabs to The Cut.)
St. Clair says she has not spoken to Musk directly about the issue, nor has she communicated with anyone directly at X. “I’m making reports through the same avenues as everyone else,” she says. She was particularly appalled by Musk’s reply to the user defending Grok, as well as the argument posed by some of Musk’s acolytes that women should avoid posting photos of themselves on the internet if they don’t want their images to be altered. “How can you possibly hold the position that everyone needs to post on X because it’s the town square,” she asks, “and also saying, ‘If you don’t want to be raped by the chatbot, log off?’”
For now, there’s little recourse for women whose photos have been altered by Grok users. On Monday, a United Kingdom watchdog agency issued a statement saying it had been in touch with X “to understand what steps they have taken to comply with their legal duties to protect users in the UK,” and French authorities are also reportedly investigating the trend. But Grok appears to continue to ignore women’s pleas to take down the edited photos. “There’s a certain pathology within that circle of people that they are above the law and they will not face consequences,” says St. Clair. “So why would X care?”
