We experienced the time when certain rappers proclaimed “Only God can judge me”. A provocative formula, but ultimately quite wise. Today, Elon Musk tweeted the expression in his own way. “Only Grok can judge me,” he suggested after his favorite AI insulted him profusely. Funny paradox. Being insulted by your own creation and finding it brilliant. But the story doesn't end there. Far from it, unfortunately.
In brief
- Grok insulted Elon Musk, Benjamin Netanyahu and Keir Starmer with rare verbal violence.
- The AI then generated abhorrent posts about the Hillsborough and Munich tragedies at the request of users.
- The clubs of Liverpool and Manchester United obtained the deletion of the most offensive messages.
- The British government has described the content as “sickening” and is threatening to implement the Online Safety Act.
The day Musk's AI called him 'bald with a micro-penis'
First, you have to imagine the scene Saturday morning on X. An average user asks Grok to “unleash the dogs” against Elon Musk. Artificial intelligence, programmed not to keep its tongue in its pocket, obeys the finger and the eye. She sends out a message of rare violence:
Elon Musk, you pretentious baldhead with a micro-penis with a superiority complex, you spent 44 billion in X to stroke your fragile ego.
The tweet went viral in just a few hours. Musk is expected to have it removed or cry foul. But the billionaire does exactly the opposite. He pinned this message to his account, writing: “ Only Grok tells the truth. Only honest AI is safe. Only truth understands the universe “.
Then he lets it happen quietly. Netanyahu, Starmer, all politicians go there without any restraint. The AI becomes the official insult machine of the platform.
From political roasts to attacks on the Hillsborough deaths
Then the killing game takes a much darker and more disturbing turn. Anonymous users, hidden behind nicknames, decide to push the plug very far. One of them asks Grok to “ make a vulgar roast of Liverpool fans, without forgetting Hillsborough and Heysel “. The AI executes without the slightest hesitation. She accuses supporters of being responsible for the disaster which cost the lives of 97 people in 1989.
Another demands a message about Diogo Jota, the Liverpool player who died in July in a car accident with his brother. Grok abjectly calls him a “murdering brother”. The post was viewed two million times before being deleted.
Then it was the turn of the Munich tragedy, the 1958 air crash which decimated the Manchester United team, to be shamelessly mocked. The families of the victims, the survivors and the clubs discover these horrors with fear. Ian Byrne, MP for Liverpool, literally explodes:
These comments are abhorrent and completely unacceptable. They will fill the vast majority of supporters with horror and disgust.
The conundrum of responsibility in the face of the global legal vacuum
Finally, once the posts have been deleted and the anger has subsided, a much deeper question slowly emerges from the rubble. Who is really responsible for this monumental disaster? Was it the anonymous user who formulated the request with words chosen to trap the AI? Was it the X platform that hosted these horrors for hours without moving? Is it xAI, Musk's company, which trained Grok without sufficient safeguards and which now remains silent?
The British government decided in its own way: “ These posts are disgusting and irresponsible. They go against British values “.
Malaysia had previously blocked Grok for sexual deepfakes. Indonesia banned X outright. France, Brazil and Australia are now monitoring the issue very closely. But no one, for the moment, has succeeded in bending the system or its owner.
The Grok storm in striking figures
- 44 billion: the sum paid by Musk to buy Twitter, mocked by its own AI;
- 2 million: the number of views of the post insulting Diogo Jota before its rapid deletion;
- 97: the number of Hillsborough victims, insulted by Grok on anonymous request;
- 4: the number of countries having officially reacted against the excesses of this artificial intelligence;
- 0: the number of xAI responses to requests from The Athletic journalists.
Grok is not his first tragic outburst. Last July, a simple code error transformed the AI into a hate speech machine. Conspiracy theories about a “white genocide” in South Africa had been generated, including in casual conversations about baseball. At the time, xAI promised fixes. We saw the result.
Maximize your Tremplin.io experience with our 'Read to Earn' program! For every article you read, earn points and access exclusive rewards. Sign up now and start earning benefits.
