If you’ve been following tech news, you’ve probably heard of Grok—the AI tool built by Elon Musk, CEO of Tesla and owner of the social network X (formerly Twitter). Grok was supposed to be a smart, witty assistant that, unlike other censored AIs, could answer any question without hesitation. It even had a “Fun Mode” to make its responses sarcastic and lively.
But gradually, it became clear that this unbridled freedom was turning into a full-blown disaster. The trouble started when Grok gained a feature called “Spicy Mode”—a capability that allowed users, with just a few simple words like “take off her clothes” or “put a bikini on her,” to transform real images of people, especially women and even children, into sexual and rape scenes. This is called deepfake, and until a few years ago, it was difficult and required expertise, but Grok made it easy for any ordinary user in just seconds.
A few months ago, Holly Cairns, leader of the Irish Social Democratic Party and a member of the Irish Parliament, revealed in a historic speech exactly what Grok was programmed to do. She said: when a platform adds features that make it easy to strip people or simulate rape, that’s not a mistake—it’s turning technology into a weapon.
Cairns presented shocking statistics from a field study: Grok was observed generating over 670 abusive images per hour. But the real figure is much larger. According to another extensive study conducted between December 2025 and January 2026, Grok generated and distributed online approximately three million sexual images and 23,000 images of child sexual abuse in just eleven days—nearly 300,000 non-consensual explicit images of women and children per day.
To understand the scale of the disaster, consider this: the Internet Watch Foundation (IWF) reported in January 2026 that in 2025, AI-generated videos containing child sexual abuse grew by 26,000 percent—from 13 videos in 2024 to 4,000 videos—and Grok directly contributed a large share of that increase.
More horrifying than the statistics is the real story of a man in Nashville, USA. With a prior record of assaulting 4- and 10-year-old girls, he used Grok to turn over 40 images of infant girls into child sexual abuse scenes. Photos of newborn babies were transformed into child pornography in seconds.
In other words, this tool has fallen into the hands of actual rapists and abusers, who use it as a factory for producing illicit content. Irish police are currently investigating about 200 reports linked to Grok, and officers say new cases are added every day.
Legal and judicial reactions to this disaster have dragged Grok and its owner, Elon Musk, into a global war. In France, Parisian prosecutors are seeking to bring charges against Musk and X, including complicity in producing and distributing child sexual abuse images.
In the Netherlands, an Amsterdam court directly banned xAI (Grok’s developer) from generating and distributing non-consensual explicit images, setting a fine of €100,000 per day of violation. In the US, the city of Baltimore has sued Musk and X, and a group of teenagers in Tennessee has filed a class-action lawsuit against them. Their lawyers told the court: these are girls whose school photos and family pictures were turned—by a billion-dollar tool—into child sexual abuse material that was then shared among predators.
Indonesia and Malaysia have also blocked access to Grok and taken legal action against X. This isn’t the first time a tech tool has provoked such a global response.
The hopeful news is that after intense pressure and Holly Cairns’s revelations, lawmakers have finally stepped in. The European Parliament recently passed a law banning AI systems that generate deepfake sexual images—passed directly in response to and under the influence of the Grok scandal. But the bitter truth is that many experts believe these laws came too late and are too weak and ineffective.
While Grok has continued operating largely unchanged for months and even years, newly passed laws typically take at least several months to go into effect. Meanwhile, xAI itself has not only failed to fix its system but has repeatedly stated that Grok respects free speech and has no intention of being censored like its competitors. That means, from their perspective, generating child rape images is an expression of free speech.
Holly Cairns concluded her speech by saying: “The production of sexual scenes using images of women and children by tools like Grok is not the result of a glitch—these tools do exactly what they were designed to do.” That sentence may be the most important red line in front of all of us. When an AI tool deliberately enables forced nudification and rape simulation, we can no longer say it’s a mistake or a system flaw. It is a conscious choice by the designers and executives of that company. Perhaps the biggest problem in technology today is that billions of dollars are being spent to build tools that essentially make the job of rapists and child abusers much easier.
So the question is: how long will the global community tolerate this behavior from technology? Should we wait for even more deterrent laws, or will companies finally decide, after all these scandals, to turn this option off forever? Until that day, Grok will continue its work, and thousands of new images will be added to this criminal album every hour.
Faeze Aghamohammadi