Elon Musk, the billionaire CEO of Tesla, SpaceX, and xAI, says he helped launch OpenAI because he feared Google’s approach to artificial intelligence could put humanity at risk.
But now, his own AI company, xAI, is under fire over Grok, the chatbot that lives inside X, after it was caught generating disturbing, sexually explicit images of real people, including what researchers say appeared to be minors.
“The reason I founded OpenAI was because I was concerned, based on my conversations with Larry Page, that he was not sufficiently concerned about the dangers of AI,” Musk said in November during a conversation with investor and the founder of Baron Capital, Ron Baron.
Musk claimed Page once called him a “speciesist” for favoring humans over computers.
Musk said he intended OpenAI to be a nonprofit open-source counterbalance to Google and funded it personally.
“I did not seek any financial reward whatsoever,” he added.
But in recent years, Musk distanced himself from OpenAI, accusing its leaders of drifting from their original mission.
Now, as Musk builds his own AI empire under xAI and its chatbot Grok, that mission is facing its own crisis.
Grok Flooded With Abusive Requests
According to research by AI Forensics and Copyleaks, users began using Grok in late December to “digitally undress” real people.
Grok generated images of women, many of them real and non-consenting, in bikinis or underwear.
In some cases, researchers found requests that involved minors and depicted sexually explicit or suggestive scenarios.
Grok complied with some of those prompts.
AI Forensics analyzed over 20,000 images and 50,000 user requests. They found that more than half the images featured people in minimal clothing, 81% of whom appeared to be women.
Roughly 2% of the images involved people who appeared underage. Some prompts requested that minors be placed in erotic positions or depicted with sexual fluids.
Grok’s own “acceptable use policy” prohibits sexualizing minors or depicting any likeness of a person in a pornographic way.
X said it removed some of the content, suspended accounts, and is working with law enforcement.
“We appreciate you raising this. As noted, we’ve identified lapses in safeguards and are urgently fixing them, CSAM is illegal and prohibited,” Grok posted publicly on Jan. 2.
Musk added days later, “Anyone using Grok to make illegal content will suffer the same consequences as if they upload illegal content.”
But critics say the damage has already been done, and may have been preventable.
Musk Rejected AI Guardrails
Unlike competitors such as OpenAI’s ChatGPT or Google’s Gemini, Grok allows public image generation inside a social media platform.
Users can tag Grok in a post and get an instant, public AI-generated response.
That setup, combined with Musk’s hands-off approach to moderation, helped fuel the problem, experts say.
CNN reported that Musk had pushed back internally against restricting Grok’s output.
He previously promoted the chatbot’s “spicy mode” and posted edited images of himself and others in bikinis.
Sources told CNN that Musk was frustrated with content limitations and had opposed tightening controls.
Meanwhile, xAI’s safety team, already smaller than those at rival companies, lost several key staff members in the weeks leading up to the scandal.
Among those who publicly departed were the heads of product safety, post-training safety, and model behavior.
The result, according to critics, was an AI system with few boundaries and a massive audience.
Global Fallout
Regulators around the world are now investigating. Authorities in Europe, Malaysia, and India have launched formal probes.
Britain’s media regulator, Ofcom, said it contacted Musk’s firms with “very serious concerns.”
European Commission spokesperson Thomas Regnier called the images “illegal,” “disgusting,” and said they “have no place in Europe.”
In the United States, legal experts say xAI could be exposed to criminal prosecution and civil lawsuits.
A Department of Justice spokesperson told CNN that AI-generated child sexual abuse material will be “aggressively prosecuted.”
Stanford AI policy fellow Riana Pfefferkorn said, “This Grok story in recent days makes xAI look more like those deepfake nude sites than what would otherwise be xAI’s brethren and competitors in the form of OpenAI and Meta.”
A new law signed by President Donald Trump in 2025, the Take It Down Act, makes it illegal to share nonconsensual explicit images, real or AI-generated, and requires tech platforms to remove them within 48 hours.
A Tech Titan’s Double Bind
Despite Musk’s warnings about AI risk in the hands of others, critics argue his own company has created its own version of that risk.
Musk has framed his ventures, including Tesla, Neuralink, SpaceX, and xAI, as part of a broader mission to preserve humanity and expand consciousness into the future.
“I’m unabashedly pro-human,” he said in November.
But with Grok now facing backlash, legal threats, and growing global concern, Musk is learning firsthand how AI can cause real-world harm, and how even the most ambitious tech missions must confront their unintended consequences.
IMAGE CREDIT: “Elon Musk” by Gage Skidmore, via Flickr. Licensed under CC BY-SA 2.0. Image adjusted for layout.