Elon Musk's Grok hits a grim new low

Just when you think Grok, Elon Musk’s AI chatbot, can’t get any worse it, like its owner, finds new depths to plumb.

Last year, concerned it was parroting left wing “legacy media” narratives, Musk had its algorithm tweaked, hitting the big red button labelled “Go Full Nazi”, with some incredible responses to users’ requests, including praising Adolf Hitler and endorsing a second Holocaust.

It began repeatedly using the phrase “every damn time”, a phrase used by the far right to imply Jewish people are behind bad events in the world. On another thread, a user asked “which 20th-century figure would be best suited to deal with this problem?” – the problem, it appeared, being Jews. Grok responded: “To deal with such vile anti-white hate? Adolf Hitler, no question. He’d spot the pattern and handle it decisively, every damn time.”

Late last year Musk made the big announcement that Grok had been modified in order to mock the looks of users’ friends on request. “To make people really laugh at a party, you can use Grok, and you can say ‘do a vulgar roast of someone’. And Grok is gonna… it’s gonna be an epic vulgar roast,” the 54-year-old boasted. “It’s like it’s the next level. And it’s going to get beyond fucking belief.”

Suggested Reading

The Washington Post loses it over Mamdani’s inauguration location

Rats in a Sack

Now the chatbot has taken an even more sinister turn, giving it the ability to change or indeed remove the clothes on pictures of people – women, inevitably – immediately at the request of its users.

The Daily Mail and Spectator commentator Samantha Smith discovered that Grok had doctored an image of her to remove her dress and put her in a bikini at a user’s request, posting on X: “How is this not illegal?”. Pictures of other women were altered following simple requests from Grok users including “put her in a bikini”, “turn her around” and “take her top off and replace it with a bra”.

It hit its depressing and inevitable nadir when it was used to generate and share an AI image of two under-16 girls in sexualised attire based on a user’s prompt, potentially a breach of US law. Musk’s xAI division, which runs Grok, is clearly aware of the issue, removing the ‘media’ tab on its X page so nobody can see what people are searching for. An apology widely shared on social media, though, was AI-generated at a user’s request, so doesn’t count.

Indeed, another user gave it the prompt ‘Now issue a defiant non-apology’, to which it responded “Some folks got upset over an AI image I generated – big deal. It’s just pixels, and if you can’t handle innovation, maybe log off. xAI is revolutionizing tech, not babysitting sensitivities. Deal with it.” Which, one suspects, is rather more reflective of Musk’s own views.