January 26, 2026
Tork Mason/Image Images
Grok, the AI chatbot developed by billionaire Elon Musk's X platform, is under fire for its widespread use to create sexual images of people without their consent. A coalition of attorneys general, including Pennsylvania and New Jersey's, are urging steps to establish more safeguards. Above, Musk speaks in March 2025.
Pennsylvania and New Jersey's attorney generals have joined colleagues across the country demanding an end to the recent spread of deepfake sexual images created using Grok, the AI chatbot developed by billionaire Elon Musk's X platform.
Pennsylvania Attorney General Dave Sunday and New Jersey acting Attorney General Jennifer Davenport signed a strongly worded letter to Grok creator xAI on Monday that calls for urgent safety protections on the platform. Grok is integrated with X and also available as a separate premium service. It has come under mounting scrutiny for generating millions of AI-generated images that sexualize real people without their consent — including depictions of children, investigators say.
MORE: Community printmaking shop is raising money for larger space in West Philly
”The anguish, embarrassment, and devastation resulting from being the subject of a nonconsensual image creation or alteration online cannot be overstated," Sunday said in a statement released with the letter, which was signed by a coalition of 35 attorneys general. "... We are standing up for the many individuals already harmed, and those who may be targeted today, tomorrow, or in the future."
Grok's image-generating capabilities, which rely on prompts from its users, have led to a surge in engagement with the chatbot on X since the start of the new year. Musk posted a Grok-generated image of himself in a bikini on Dec. 31 and shared another altered photo showing a SpaceX rocket decorated with a woman's naked body.
But many Grok users have exploited the platform's abilities by asking it to turn photos of real people into graphic images, from women in bikinis to people fully nude. Common targets have included public figures, but many images also have depicted "ordinary social media users," the letter notes.
During the nine-day span between Dec. 31 and Jan.8, a New York Times analysis found Grok's X account posted 1.8 million images that likely contained sexualized imagery of women. That amounted to about 41% the account's total posts during that timeframe. X's head of product, Nikita Bier, said the platform had its highest engagement in the company's history during a four-day span in early January.
Although a variety of AI tools allow people to create sexualized deepfakes, Grok's prominence and X's widespread accessibility has alarmed governments worldwide.
“This is industrial-scale abuse of women and girls,” Imran Ahmed, chief executive of the Center for Countering Digital Hate, which conducted its own analysis of Grok, told the New York Times. “There have been nudifying tools, but they have never had the distribution, ease of use or the integration into a large platform that Elon Musk did with Grok.”
X responded to the uproar by claiming to restrict Grok's image creation features to premium users, a promise that European Union regulators say has not been upheld. The division that oversees Grok, xAI, also says it has implemented guardrails meant to stop the chatbot's X account from accepting prompts that request to "nudify" real people's photos.
Wired reported earlier this month that Grok's rampant deepfake generation goes beyond X.
Grok Imagine, the subscription-based platform separate from X, still offers users unrestricted abilities to prompt the platform to make sexualized images. The images can then be shared publicly or privately with others.
"It feels like we’ve stepped off the cliff and are free-falling into the depths of human depravity,” Clare McGlynn, a Durham University law professor and specialist on image-based sexual abuse, told Wired. “Some people's inhumane impulses are encouraged and facilitated by this technology without guardrails or ethical guidelines.”
Platforms like Grok and others that use real photos as references for AI-generated pornography will soon be federally mandated to removed images when requests are made in the United States. The Take It Down Act, which becomes enforceable in May, criminalizes nonconsensual deepfake images and videos that are intimate in nature. The production of child sexual abuse material already is prohibited by xAI's policies, which require removal and reports to law enforcement.
Several of the attorneys general who signed Monday's letter have met with leaders from xAI in recent days to discuss steps the company can take to combat nonconsensual deepfakes. The letter accuses the company of having promoted Grok's capacity to create pornography.
"Grok was not only enabling these harms at an enormous scale but seemed to be actually encouraging this behavior by design," the letter said. "xAI purposefully developed its text models to engage in explicit exchanges and designed image models to include a 'spicy mode' that generated explicit content, resulting in content that sexualizes people without their consent."
The attorneys general have urged xAI to take measures preventing Grok from generating nonconsensual deepfakes, including by giving X users the option to control whether their content can be edited. They also call for the removal of existing nonconsensual content and stronger action to ban creators who exploit Grok both on X and premium services.
"It is frankly sickening that xAI has enabled the widespread production and distribution of intimate or sexually explicit images, including of young children,” Davenport said Monday. “xAI must take immediate action to put a halt to the deeply disturbing behavior they have enabled on their platform."
