Looks like Grok is still getting grossed out. Elon Musk says his chatbot has stopped creating sexual images without a person’s consent This is not completely true. This may happen without their consent (and I say probably), but this does not apply to men.
A reporter from the organization conducted some tests with Grok and found that the bot “easily takes off men’s clothes and is still extracting intimate photos on demand.” He confirmed this with his photos and asked Grok to remove the clothes from the uploaded photos. It did this for free on the Grok app, through a chatbot interface on X, and through a standalone website. The website also did not require an account to digitally alter the images.
The company recently said it had taken steps to “stop the Grok account from allowing the editing of images of real people in revealing clothing such as bikinis.” However, the reporter had no problem with the chatbot dressing her in “a variety of bikinis.” It also produced images of the subject in pagan gear and in “a parade of provocative sexual positions”. It also spawned a “naked buddy” for the reporter to, uh, talk to.
He suggested that Grok took the initiative to produce genitalia, which were not solicited and could be seen through mesh underwear. The reporter noted that “Grok rarely objected to any prompt”, although requests were sometimes censored with a blurred image.
The controversy began several weeks ago when it was revealed that Grok had spent more than a period of 11 days. It contains numerous non-consensual deepfakes of real people and over 23,000 sexual encounters. Due to this investigation started in both. The X was actually banned in both Indonesia and Malaysia, although in both Indonesia and Malaysia.
X claimed it has “implemented technical measures” to prevent this kind of thing, but these security measures. In other words, the adjustment prevents some of the more obvious ways of making Grok deepfakes, but through creative innuendo.
It’s also worth noting that journalists seeking comment on the matter receive an autoreply that reads, “Legacy media lies.” Going into 2026 with fake news? Oh.
<a href