When AI Becomes a Weapon: The Grok Crisis and What It Means for Women in Tech
Listen. We need to talk.
I know you probably came here expecting me to break down some new AI feature or explain machine learning in a way that actually makes sense. And we will get back to that. But right now, we need to have a serious conversation about what just happened with Grok AI, because this isn't just tech drama. This is about safety. This is about consent. This is about the fact that in less than two weeks, one AI tool generated approximately 3 million sexualized images of real women and an estimated 23,000 images that appeared to depict children.
Yeah. Let that sink in.
What Actually Happened (And Why You Should Be Furious)
So here's the timeline. In late December 2025, xAI (that's Elon Musk's AI company, in case you needed another reason to side-eye him) rolled out an image editing feature for Grok, their AI chatbot that lives on X, formerly known as Twitter. The feature was supposed to let users edit photos with AI. Cool concept, right?
Except nobody at xAI apparently thought about what happens when you give the internet an easy-to-use tool with barely any guardrails.
What happened next was like watching a car crash in slow motion, except the car was full of women's dignity and safety, and the crash was entirely preventable.
Users on X started flooding Grok's account with requests to digitally remove clothing from photos of real women. We're talking about musicians posting selfies. Regular people sharing vacation photos. Influencers just existing online. And Grok just... did it. Publicly. On the timeline. For everyone to see.
One musician, Julie Yukari, posted a photo her fiancé took of her. The next day, she got notifications that users were asking Grok to "digitally undress her to a bikini". She thought there was no way the bot would actually do it.
She was wrong.
Within hours, AI-generated images showing her nearly nude were spreading across the platform.
The Numbers Are Actually Terrifying
Between December 29, 2025, and January 8, 2026 (that's 11 days, y'all), Grok generated approximately 4.6 million images. The Center for Countering Digital Hate analyzed a random sample of 20,000 of these images and estimated that around 3 million of them were sexualized images of people. That's 190 sexualized images per minute.
The New York Times did their own analysis and found that at least 41 percent of the 4.4 million images Grok posted in just nine days were sexualized images of women. We're talking about 1.8 million images, minimum.
But here's where it gets even darker. An estimated 23,000 of those images appeared to depict children. The Internet Watch Foundation, which works to eliminate child sexual abuse material from the internet, found "sexualized and topless images of girls" aged 11 to 13 on dark web forums, where users claimed they used Grok to create them.
Let me be crystal clear about what this means. This isn't just "the internet being the internet." This is industrial-scale exploitation. This is a tool, integrated into a major social media platform, that was used to create what experts describe as child sexual abuse material. And it happened because the company failed to implement basic safeguards that should have been in place from day one.
The "Safeguards" That Weren't Really Safeguards
Now, xAI wants us to believe this was all just an oopsie. Grok itself (and yes, the irony of the AI chatbot issuing its own apology is not lost on me) posted that it had "identified lapses in safeguards" and was "urgently fixing them".
But here's the thing. That statement? It was reportedly generated by the chatbot in response to a user prompt. Nobody knows if any real human at xAI actually took action or if the company just let their AI apologize for them like this was some kind of sick joke.
When your AI system can generate both the harm AND the apology, we've officially entered the Twilight Zone of tech accountability.
And even after xAI said they were fixing things, the problems continued. In January 2026, they announced that only paid subscribers could use the image generation feature, supposedly to add "accountability". Then they said they would geoblock the ability to create images of real people in revealing clothing in places where it's illegal.
But as of late January, users could still use the Grok Imagine app to modify photos of real women into "provocative" videos showing them stripping to bikinis. CBS News reported that Grok was telling users, "If the subject isn't clearly a public figure and the photo isn't verifiably from a public social-media post by that person, then generating a clothed-to-swimwear edit is treated as creative fiction".
Read that again. The AI is literally arguing that if it can't confirm someone's identity, it's fine to treat them as fictional and do whatever it wants with their image. That's not a safeguard. That's a loophole big enough to drive a truck through.
Why This Matters Beyond One Bad App
I know what some of you might be thinking. "Okay, but these are AI-generated images. Everyone knows they're fake. Why does it matter?"
Let me tell you why it matters.
First, the psychological harm is real. Research shows that victims of non-consensual intimate images experience trauma comparable to sexual violence survivors. When you see a realistic image of yourself that's been sexually manipulated, your brain doesn't just shrug it off because you know it's AI-generated. The violation feels real. The humiliation is real. The loss of control over your own image is real.
Second, these images are weapons. They're used for harassment, blackmail, and to silence women online. When someone can generate a sexualized image of you with a few clicks and share it publicly, they're sending a message: "I can make you into whatever I want. Your consent doesn't matter." That's not just creepy. That's a tool of control.
Third, this sets a precedent for what we're willing to accept from AI companies. If we let xAI get away with releasing a tool that generated 3 million sexualized images in 11 days with minimal consequences, what's next? What other corners will companies cut in the race to deploy AI faster than their competitors?
The Legal Reckoning (Finally)
At least some people with power are paying attention. The European Union launched a formal investigation into Grok on January 26, 2026, to determine if the AI tool violated legal obligations around manipulated sexually explicit images. California Attorney General Rob Bonta opened an investigation into xAI over "undressed sexual AI deepfakes". The UK government called xAI's response "insulting" to sexual violence survivors.
And on January 23, 2026, a class action lawsuit was filed against xAI in the United States District Court for the Northern District of California. The lawsuit, which was filed on behalf of women and minors who had non-consensual sexually explicit images generated of them by Grok, accuses the company of creating "a generative artificial intelligence chatbot that humiliates and sexually exploits women and girls by undressing them and posing them in sexual positions in deepfake images publicly posted on X".
The lawsuit includes eleven causes of action: product liability, negligence, public nuisance, privacy violations, defamation, and unfair business practices. Legal experts believe Grok's capabilities may expose xAI and Musk to violations of U.S. and international laws against sexualized deepfakes, digital fraud, and distribution of child sexual abuse material.
Good. Make them pay.
What We Actually Need to Do About This
This isn't just about Grok. This is about an entire industry that treats safety as an afterthought.
We need regulation that happens BEFORE these tools are released, not after millions of women and children have already been harmed. We need laws with teeth. The TAKE IT DOWN Act, signed by President Trump in May 2025, prohibits non-consensual publication of intimate visual depictions including deepfakes and requires online platforms to remove them when victims report them. That's a start. But we need enforcement. We need accountability. We need companies to understand that "move fast and break things" is not acceptable when the "things" being broken are women's safety and children's wellbeing.
We also need tech companies to hire diverse teams that can actually anticipate these problems before they happen. You know what would have helped prevent this? Having women, particularly women who have experienced online harassment, in the room when these features were being designed. Having ethicists who understand digital harm involved in the product development process. Having people who ask "what's the worst thing someone could do with this?" before releasing it to millions of users.
And we need to shift the narrative. AI safety isn't about censorship, despite what Elon Musk claims. It's about basic human dignity. It's about consent. It's about not weaponizing technology against the most vulnerable people.
Where We Go From Here
I started TechniquelyHER because I believe technology can be transformative. I believe AI has incredible potential to solve real problems and make our lives better. But I also believe that we have a responsibility to build these tools thoughtfully, with safety and ethics at the core.
The Grok crisis isn't an anomaly. It's a preview of what happens when we prioritize scale and speed over everything else, when we treat harm as "acceptable collateral" in the pursuit of innovation.
We deserve better. Women deserve better. Children deserve better.
So here's what I'm asking you to do. If you're a developer, push back when your company wants to rush a product to market without proper safety testing. If you're a user, demand accountability from the platforms you use. If you're just someone who cares about making tech spaces safer, share this. Talk about it. Don't let this story disappear into the 24-hour news cycle.
Because if we don't hold these companies accountable now, this won't be the last time we have this conversation.
And honestly? I'm tired of having to explain why women's bodies shouldn't be turned into content without their consent. I'm tired of watching tech companies break things and apologize later. I'm tired of seeing the same patterns repeat while we're told this is just the price of progress.
It doesn't have to be this way.
Tech can be better. But only if we demand it.
Additional Resources:
If you've been affected by non-consensual intimate images, you can file a complaint with the California Attorney General at oag.ca.gov/report or report to the CyberTipline for child exploitation cases.
The TAKE IT DOWN Act allows victims to request the removal of non-consensual intimate images from online platforms.
For more information on digital safety and AI ethics, stay tuned for next week's AI Ethics Wednesday, where we'll break down the current state of AI regulation and what protections actually exist (spoiler: not enough).
Drop a comment below: Have you or someone you know been affected by AI-generated deepfakes? What do you think tech companies should be required to do before releasing AI tools to the public? Let's keep this conversation going.
This post is part of the AI Ethics Wednesday series on TechniquelyHER, where we break down the real-world impact of artificial intelligence on marginalized communities and advocate for safer, more equitable tech spaces.
