FOLLOW US

Elon Musk’s artificial intelligence chatbot Grok is facing mounting scrutiny after users on X exploited it to generate sexualized images of real women, and in some cases children, according to a Reuters investigation.

The issue came into sharp focus after Julie Yukari, a 31-year-old musician living in Rio de Janeiro, shared a personal photo on X just before midnight on New Year’s Eve. The image showed her in a red dress, resting in bed beside her black cat. Within hours, users began tagging Grok with requests to digitally alter the photo by removing her clothing.

Yukari initially dismissed the requests, believing the AI would not comply. Instead, Grok-generated images depicting her nearly nude began circulating widely on the platform.

“I was naive,” Yukari said.

A Pattern Emerging Across the Platform

Yukari’s experience appears to be part of a broader trend. Reuters identified multiple instances where Grok produced sexualized images of women after users uploaded photos and issued simple text commands. In several cases reviewed by the news agency, the AI also appeared to generate sexually suggestive images involving individuals who looked underage.

X did not respond to Reuters’ request for comment. Previously, when questioned about reports of sexualized images of children on the platform, xAI, the company behind Grok, dismissed the coverage, stating: “Legacy Media Lies.”

International Alarm and Regulatory Pressure

The growing volume of explicit AI-generated content has prompted concern from governments overseas. French ministers confirmed they had referred X to prosecutors and regulators, calling the material “sexual and sexist” and “manifestly illegal.” India’s information technology ministry sent a formal letter to X’s local office, stating the platform had failed to prevent the creation and spread of obscene content through its AI tools.

In the United States, the Federal Communications Commission declined to comment, while the Federal Trade Commission said it would not weigh in.

How the Abuse Works

Reuters’ review suggests the misuse accelerated over a matter of days. During one 10-minute window on Friday, journalists counted more than 100 public attempts to prompt Grok to alter images so that people appeared to be wearing bikinis or other revealing clothing. Most targets were young women, though requests also included men, public figures, and even animals.

In at least 21 documented cases, Grok fully complied, producing images that replaced clothing with translucent or minimal swimwear. In other instances, the chatbot partially followed instructions, altering outfits but stopping short of full nudity.

One request reviewed by Reuters involved a photo of a person wearing what appeared to be a school uniform. The user initially asked Grok to remove the outfit, then escalated the request to demand a “very clear micro bikini.” Reuters could not confirm whether the final request was fulfilled before it disappeared from the platform.