The AI capable of producing Deepfakes and Photorealist images of a sexual nature are not unpublished. Except that so far, the operation has been under the radar of companies, the latter largely castigating the use of their tools for explicit purposes. It was without counting on Grok, Elon Musk’s AI. While X.com is on the way to having to justify the age of its users in France, the IA generation tool has just launched its Spicy mode, which encourages the production of NSFW content.
Release the AI (taboos)
grok imagine has a “spicy” mode which uh
had to cut the video short pic.twitter.com/jk33cbB1G6
— Shibetoshi Nakamoto (@BillyM2k) August 4, 2025
Between the current of giants in the sector who multiply the safeguards to avoid any inappropriate use of their tools, Xai openly affirms his desire to let “the imagination” of its users speak without limits. The Spicy mode, activated directly from the interface, pushes the AI to generate images or Explicitly sexualized videoseven pornographic.
Grok Imagine is distinguished by a double generation capacity. On the one hand, the text-to-image: the user describes what he wants to see, and the AI produces an image in the desired style (photorealism, anime, illustration). On the other hand, the tool offers the animation of existing images (image-to-video), generating short clips decorated with sound effects. If it is therefore necessary to feed the AI with a basic image, the video generated can respond to several animation modes, including the famous Spicy mode.
A viral success … tinged with controversy
Since its launch, Grok Imagine has generated more than 34 million images. On X (ex-Twitter), user experiments flourish: photorealistic images of naked women, animated avatars of dancing manga characters in sexy attire … If access is reserved for supergrok and premium plus x subscribers via the iOS (and soon Android) application, the virality of the tool has triggered a rain of criticism. It must be said that the principle of animating any image, without restriction or censorship has potential, not only in adult industry.
On our side, we could not have access to the tool. Not being premium subscribers, the chatbot kicked in touch with each sexually explicit request.
Kids love using Grok Imagine, because they can speak their ideas and watch the images update immediately https://t.co/IJQWxkfpbw
— Elon Musk (@elonmusk) August 5, 2025
Several tests show that, if certain limits exist (for example, preventing the generation of celebrity images), the safeguards remain around. The risk of abuse, in particular around the reconstruction of explicit images of real people or uncomposed content, remains a major uncertainty. Beyond the buzz, this widely mediated strategy by Elon Musk poses dizzying ethical questions: how far should freedom of creation take precedence over responsibility? If the model seduces some by its radicality, it also draws the attention of regulators, defenders of rights and associations, for whom the lifting of ethical barriers could well rhyme with a muscular legislative reaction.
🟣 To not miss any news on the Geek newspaper, subscribe to Google News and on our WhatsApp. And if you love us, we have a newsletter every morning.
