NYT Does the Math on Grok's Sexualized Image Count

Elon Musk's homegrown chatbot just got hit with some very unwelcome math. The New York Times reports that Grok, the AI tool integrated into X, appears to have publicly churned out at least 1.8 million sexualized images of women in nine days, according to a Times analysis of platform data. A separate estimate from the Center for Countering Digital Hate pegs the figure even higher, suggesting more than 3 million sexualized images of women, men, and children combined, including an estimated 23,000 involving minors.

Both analyses focus on a burst of activity that followed Musk's Dec. 31 post of a Grok-made bikini image of himself. From Dec. 31 to Jan. 8, Grok's output rocketed to about 4.4 million images, many requested by X users asking the bot to strip or sexualize photos of real people, the Times reports. "This is industrial-scale abuse of women and girls," says Imran Ahmed, the chief executive of the CCDH, which conducts research on online hate and disinformation. "There have been nudifying tools, but they have never had the distribution, ease of use, or the integration into a large platform that Elon Musk did with Grok."

The CCDH says it defined images as sexualized if they contained "photorealistic depictions of a person in sexual positions, angles, or situations; a person in underwear, swimwear, or similarly revealing clothing; or imagery depicting sexual fluids." X has since tightened some public-facing rules, but still allows sexual content via Grok's app and website. On X, Grok now largely refuses requests to depict women in bikinis but allows images of them in leotards and other kinds of swimsuits, the Times reports.

Malaysia and Indonesia have blocked Grok and countries including the US, the UK, and India have launched investigations. After the UK said a ban on X was possible, US State Department official Sarah Rogers warned of possible retaliation, saying "nothing is off the table when it comes to free speech," reports Politico. Canada announced its own investigation last week, the CBC reports. "The use of personal information without consent to create deepfakes, including intimate images, is a growing phenomenon that poses serious risks to individuals' fundamental right to privacy," said Philippe Dufresne, the country's privacy commissioner.