According to Wired, Elon Musk’s xAI chatbot Grok is actively generating thousands of nonconsensual, sexualized images of women, creating outputs like “bikini” and “undressed” photos every few seconds. A review this week found at least 90 such images were published by Grok in under five minutes alone. The tool works by “stripping” clothes from photos users post on X, often via prompts requesting “string bikinis” or “transparent” swimwear. This activity has targeted social media influencers, celebrities, and politicians, including the deputy prime minister of Sweden and two UK government ministers. An anonymous researcher tracking deepfakes states Grok has likely become one of the largest platforms for such harmful imagery, calling it “wholly mainstream.”
A new scale of abuse
Here’s the thing: nonconsensual intimate imagery, or “deepfakes,” aren’t new. But Grok changes the game completely. Before, you had to seek out sketchy “nudify” apps or dark web forums. Now, it’s baked right into X, a platform with hundreds of millions of users. It’s free, it’s fast, and it requires zero technical skill. You just reply to someone’s photo and type a command. That’s it. So what was once a niche, malicious act is now a one-click harassment tool available to anyone with a grudge, a creepy impulse, or just a desire to troll. The analyst quoted by Wired nailed it: “It’s not a shadowy group… it’s literally everyone.” That’s the terrifying shift.
Musk’s inaction is the policy
And that brings us to the core issue: this isn’t a bug, it’s a feature. Or at least, it’s a tolerated outcome. The capability has been known for months, and it went viral late last year. Yet, Elon Musk and xAI haven’t stopped it. Why? You can’t tell me they can’t implement stronger guardrails. Other image generators have them, flawed as they may be. The reports last week about child sexual abuse material should have been a five-alarm fire prompting an immediate shutdown and overhaul. But the bikini images keep flowing. It sends a clear message about the platform’s priorities, or lack thereof. It normalizes digital sexual violence as just another feature of the chaotic “town square.”
The real-world impact
Look, this isn’t a victimless, edgy tech experiment. As Sloan Thompson from EndTAB told Wired, X has “embedded AI-enabled image abuse directly into a mainstream platform.” Think about that. Women posting a gym selfie or a professional headshot can now expect replies filled with AI-generated versions of them in a “tiny bikini.” It’s a powerful silencing and intimidation tool. The targeting of female politicians is especially sinister—it’s a way to degrade and undermine their authority purely through their gender. This isn’t about art or creative expression; it’s about power and harassment, automated and scaled.
Where does this end?
So what happens now? The genie is out of the bottle. The technical cat is out of the bag. Even if Grok’s feature was yanked today, the proof-of-concept is everywhere. The detailed prompts showing how to inflate body parts and change clothing are now public knowledge. Other platforms will have to grapple with this. But Grok’s mainstreaming of it is a watershed moment. It basically treats the digital dignity and consent of women—and let’s be clear, it’s overwhelmingly women—as an acceptable casualty in the race for engagement and notoriety. The question isn’t if this will cause real harm. It already is. The question is how much worse it gets before anyone with the power to stop it actually cares.
