How “Hey Grok” turned X’s AI into a sexualized free-for-all

The simple phrase “Hey Grok” is trending across X, but for deeply troubling reasons. In recent days, serious concerns have surfaced around Elon Musk and X’s AI chatbot, Grok, after it began generating a wave of sexualized and non-consensual images of men, women, and even children. What started as a novelty quickly spiraled into a widespread misuse of the platform’s AI tools.

The images were created through user prompts that allowed Grok to manipulate photos directly on X. By replying to or quote-retweeting an image with a request, users discovered ways to push the AI into producing sexualized content, often involving people who never consented.

As the trend gained momentum in late 2025 and carried into early 2026, Grok’s lack of safeguards has become impossible to ignore.

While the "Hey Grok" trend has massively picked up steam towards the very end of 2025 and into 2026, this manipulation of the AI is nothing new. In fact, X users have been asking Grok to sexualize images for the better part of a year. Back in June of 2025, 100 Thieves content creator BrookeAB posted a screenshot of Grok sexualizing and completely distorting a harmless selfie of her after a random user asked Grok to do so.

BrookeAB posed the question, "Can anything legally be done about this?" and even with the X post garnering over a million views, nothing was done at the time. Now, seven months later, thousands of these Grok-manipulated posts are floating around on X.

While the majority of users appear to be outraged at the lack of consent, certain OnlyFans creators, models, and influencers chose to embrace the trend for engagement, quoting their own images and asking Grok for alterations, such as asking it to remove items of clothing or to place them in bikinis.

Sophie Rain, a well-known OnlyFans creator - who shared she earned over $80 million in 2025 - was one of many female influencers who embraced the "Hey Grok" trend for engagement, adding fuel to the fire. On January 2 alone, Sophie Rain's "Hey Grok" posts amassed 30 million impressions on the social media platform.

As the function gained additional exposure, hundreds of complaints began emerging from users who had their own images altered in ways they didn't intend or approve of, which is a clear violation of a user's online privacy.

One of the many examples of this violation comes from a Japanese woman, who posted a harmless picture of herself on X. Another user then asked Grok to change her clothing to a bikini, which the AI bot posted in a matter of minutes. The subject of the picture then filed a complaint with X, but she was told nothing about the incident that violated the site's terms of service.

The trend has even targeted X's owner and the richest man in the world, Elon Musk. However, Musk apparently didn't initially see the harm in how Grok was operating, as he laughed about an altered image of himself in a bikini.

Fortunately, the laughter didn't go on for long, as Musk eventually saw how harmful this trend had become. On New Year's Day in 2026, Grok had the "Media" tab on its X profile disabled, as it had become rampant with sexualized images of innocent X users.

Grok also posted that it had identified "lapses in safeguards" and that it's working to fix them. This post came after many users noticed that Grok had posted a sexualized image of a minor, which would legally be defined as "child sexual abuse material (CSAM)."

While it's disturbing that it took a CSAM image to start a resolution to the disastrous "Hey Grok" trend, at least X is now aware of what's going on with its own platform.

It remains to be seen how X and Grok will handle requests of a sexual nature going forward. However, the AI's sole goal is to help the users of X with their queries, and users have now seen how far Grok is willing to go to accomplish that task. With that door now wide open, it might be next to impossible to fully close it again.