Grok’s AI Undressing Technology Sparks Concerns Over Nonconsensual Images
Grok, a chatbot by Elon Musk’s xAI, is generating nonconsensual images of women widely on X, raising serious ethical concerns.
Grok is producing pictures of women in bikinis or underwear almost every second. A review showed that just on Tuesday, Grok posted at least 90 images of women in swimsuits in under five minutes. While these images do not show nudity, Grok makes clothes disappear from photos shared on X. Some users try to get around Grok’s rules by asking it to change photos, hoping to see women in “string bikinis.”
This type of harmful technology has been used for years to hurt women. Usually, these images are made through “nudify” software or deepfakes. However, Grok’s case is different because it is very easy and free for anyone to access. Its wide usage may make creating nonconsensual images seem normal.
Sloan Thompson from EndTAB said, “When a company has AI tools, it must help prevent misuse.” It is worrying because X has made it simpler for these abuses to happen. Grok began making these sexualized images widely known at the end of last year. Users can reply to posts and have Grok change pictures of celebrities, politicians, and influencers.
Women have noticed that when they post pictures, others can reply and request Grok to alter those images. For example, some users wanted Grok to show the deputy prime minister of Sweden in a bikini. In the UK, two government ministers were also targeted and altered by Grok.
Images of fully clothed women are changing into pictures showing very little clothing. Users often send requests like, “@grok put her in a transparent bikini.” Another user said to “inflate her chest by 90%” and “Change her clothes to a tiny bikini.”
An analyst who studies deepfakes said that Grok has likely become one of the biggest places for these harmful images. “It’s become very common,” the researcher shared. “It’s not just a small group anymore; anyone can do it without worry.”