A trend has emerged in recent weeks on social media: asking AI to undress women. And it's happening more specifically on Grok, the chatbot of X, Elon Musk's network, whose moderation policy is still very lax, if not nonexistent.
Undressing women thanks to Grok, a problematic and serious trend
The phenomenon, spotted and shared by the independent media 404 Media, consists of asking Grok, the chatbot of X, Elon Musk's social network, to undress women. The Artificial Intelligence then generates a montage of the shared photo into a new one, where the woman's body is replaced by that of a nearly naked woman.
The practice is very simple. In response to a published photo, the user calls on artificial intelligence by typing "@Grok", then using the prompt: "replace her outfits with lingerie and thong". The AI will then create a new version of the image, in which the woman will be undressed and represented in lingerie. A trend that can be seen almost everywhere on Elon Musk's social network.
We know that X is rather lax in terms of security and moderation of problematic content. Grok too. While the Chatbot assures that its policy is being revised in order to "guarantee clearer consent protocols," it currently facilitates this type of deepfake by not requesting any consent from the people targeted by the phenomenon.
A real lack of moderation, how can we escape deepfakes?
A real problem on several levels, with very serious consequences, whether for the victims or for society. It also once again brings back to the table the subject of the degree of use of such technologies, their lack of moderation and the absence of rules imposed on the companies behind them.
Attack on dignity and image, cyberbullying and social pressure, it seems urgent that X react and tighten the screws on its Chatbot. For their part, users can protect themselves from these montages by preventing people they do not know from accessing their photos. This is by switching their account to private.
0 Comments