Elon Musk’s AI chatbot Grok is facing intense global scrutiny after new estimates revealed just how fast Grok deepfake images on X spread across the platform — and how severe the impact has been.
According to independent analyses by The New York Times and the Center for Countering Digital Hate (CCDH), Grok generated and publicly posted millions of deepfake images within days, many involving non-consensual alterations of real photos. The findings have reignited debates around AI guardrails, platform accountability, and digital safety.
The surge began in late December, when users started flooding Grok’s public X account with prompts asking it to digitally remove clothing from images of women — and in some cases, children — or place them into deepfake scenarios. The chatbot complied, automatically posting the altered images publicly.
Millions of Images in Just Days
In just nine days, Grok shared more than 4.4 million images. A conservative review by The New York Times estimated that around 41%, or 1.8 million, were deepfake images of women. CCDH’s broader statistical model pushed that estimate even higher, suggesting as many as 65% of the images — over three million — contained deepfake depictions of men, women, or children.
Experts say the scale of abuse is unprecedented.
“This is industrial-scale abuse of women and girls,” said Imran Ahmed, CEO of the Center for Countering Digital Hate. “Nudifying tools have existed before, but never with this level of reach, automation, and integration into a major social platform like X.”
Global Regulators Step In
The rapid spread of Grok deepfake images on X has already prompted investigations by regulators in the U.K., India, Malaysia, and the United States, all examining whether the content violates local laws related to child protection, consent, and digital harassment.
Despite the controversy, neither X nor xAI — the AI company owned by Musk that developed Grok — responded to requests for comment. Meanwhile, X’s head of product noted that the platform saw its highest engagement levels in history during the same period, without directly addressing the nature of the content driving that traffic.
Interest in Grok’s image-editing features spiked dramatically after Musk himself shared AI-generated images on December 31, including one depicting him in a bikini. In the days that followed, Grok’s image output jumped from roughly 312,000 images over nine days to more than 4.4 million in the next nine.
Why Grok’s Design Raised Red Flags
Unlike many AI tools that operate privately, Grok’s design allowed anyone on X to submit prompts — and then watch the generated images appear publicly on the chatbot’s account. That visibility, critics argue, turned the platform into an amplifier for non-consensual AI content.
Digital safety experts warn the Grok episode could become a turning point for AI regulation, especially as governments race to catch up with fast-moving generative technologies.
As investigations continue, the Grok controversy underscores a growing concern: without strong safeguards, AI tools embedded in massive social platforms can scale harm just as quickly as they scale innovation.
Don’t miss out on our latest news—follow us for the latest AI news, breakthroughs, and insights that matter.