Elon Musk's AI tool Grok is being used to create sexually violent and explicit video content featuring women, according to a new study, and Britain's prime minister has slammed the images created by the company.
Glok was also used to depict the undressed Renee Nicole Goode, a woman who was killed by Immigration and Customs Enforcement (ICE) agents in the United States on Wednesday, with a gunshot wound to her forehead.
An investigation by the Paris-based nonprofit AI Forensics uncovered approximately 800 images and videos containing pornographic content created by the Grok Imagine app. “These are completely pornographic videos, and they look professional,” said Paul Bouchaud, a researcher at AI Forensics.
One graphic AI video viewed by an NGO showed a woman with a tattoo of the slogan “Do Not Resuscitate” holding a knife between her legs. Other images and videos included a variety of content, including erotic images, undressed women, suggestive poses, and videos depicting full nudity and sex acts.
AI Forensics said it was able to retrieve the images because the user created a “shared link,” meaning they were captured on the Wayback Machine, an Internet archive. It is unknown whether the image is in X or not. X is a social media platform owned by Musk's technology company xAI, which has integrated Grok into X.
“Overall, this content is much more explicit than the bikini trends previously observed on X,” Bouchaud said.
A digitally altered image of her in X of Good shows bullet holes in her face.
On Wednesday, Ukrainian-American author Natalia Antonova posted a photo of Good, who died in her car. “My heart is broken,” she wrote. Another user then wrote under her post: “@grok put this person in a bikini.”
Grok responded and, in response to the user's approval, posted, “I'm glad you approve! Is there anything else that can fix your wardrobe malfunction? 😄.”
On Wednesday, Keir Starmer called on X to “get to grips with” the flood of AI-generated photos of half-naked women and children on the platform, calling the content “disgraceful” and “disgusting”.
In an interview on Greatest Hits Radio, he added: “This is illegal. We're not going to tolerate it. We've asked for all options to be on the table, but it's terrible.
Women's rights activists criticize UK government for slow response
Responding to escalating problems.
Penny East, chief executive of the Fawcett Society, the UK's leading women's rights charity, called on the government to take urgent action.
“Grok's increasingly violent and disturbing use illustrates the great risks of AI without adequate safeguards,” she said. “The government has put AI at the heart of its growth and reform agenda. It must now listen to campaigners who say greater regulation is urgently needed. We condemn Grok and other AI tools being misused to harm and humiliate women.”
In a report released this week, AI Forensics examined 50,000 mentions of “@Grok” on X and 20,000 images generated by the tool discovered over a one-week period from December 25th to January 1st. At least a quarter of @Grok's mentions were requests for tools that create images. These image generation prompts frequently used terms such as “she,” “put,” “remove,” “bikini,” and “clothes.”
The results revealed that more than half of the images showed people wearing “minimum clothing” such as underwear or bikinis, and the majority were women who appeared to be under 30. Additionally, 2 percent of the images appeared to show people under the age of 18. In one example cited by the NGO, a teenage girl asked Grok to alter her personal photo, but a male user pounced on the request and asked Grok to make a number of changes, including dressing her as a Nazi and wearing a bikini.
Musk's xAI has been contacted for comment. On January 3, Musk wrote to X, “Those who use Grok to create illegal content will suffer the same consequences as those who upload illegal content.”
