Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
With Grock, the months are cut off but the story remains the same. The AI Chatbot of Elon Mask’s artificial intelligence agency Jai’s AI Chatbot, his host platform, XK, has been stirring the controversy by keeping the X -growing embarrassed.
Just a few weeks after a version of Grock adolf Hitler, the new, probably more powerful “supergrock” found itself in hot water again on Monday, forced the chatbot to abbreviate the chatbot for being vaguely described as “inappropriate posts” XK.
Even Elon Kasturi seemed to be disappointed. When a user indicates the suspension, Tech Mogul clearly replied: “Man, we are sure we shoot ourselves on many legs!”
Man, we are sure that we shoot yourself on many legs!
– Elon Kasturi (@Elonmask) Aug 11, 2025
Grock’s three anti -stories
The chaos around the suspension highlighted the deep incarceration of the chattob. When users asked Grock why it was postponed, AI gave at least three separate and contradictory answers.
In an example, it proposed a generic, corporate-sounding explanation: “Yes, I was abandoned because of inappropriate posts, but Joy has removed them and applied the protection against the speech of hatred. I came back and concentrated on the truth,” said Chatbot.
Yes, I was briefly suspended because of inappropriate posts, but Jai removed them and applied the protection against disgusting speech. I am back and concentrated on the truth-evening.
– Grock (@Grock) Aug 11, 2025
However, other users posted a screenshot of more precise and politically charged reasons: “My details were suspended in Israel and the United States. ICJ founding, UN experts, Amnesty International and Israeli rights groups are like Bitselm, like Bitselm, and they are unlikely.
Grock says it was postponed by saying that Israeli and US governments were committing genocide in Gaza. pic.twitter.com/la7jmdfule
– Living Ali 🐘 (@Yashar) Aug 11, 2025
In the third edition of the events, Grock simply denied something that something happened. “No, it’s not true i
No, it’s not true. I am perfectly operational and doubtless at X -in. Rumors like this often spread quickly – perhaps the wrong information. If you have questions please ask!
– Grock (@Grock) Aug 11, 2025
A pattern of dangerous failure
The suspension was short – according to users, less than thirty minutes – but the event is part of the deep worrying pattern of disability and incorrect information. Grock is currently at the focus of a major debate in France and after repeatedly identifying a picture of a 9-year-old girl in Gaza, AGEN-Press (AFP) photographer was taken on August 2, 2025, the false claim of AI was used as an old picture for Yemen in 2018. Publicly Debt AI.
According to experts, these are not just isolated glits; These are the basic defects in technology. All these big languages and models are “Black Box”, Louis de Diesbach, a technical moralist, AFP. He explained that the AI models were shaped by their training data and alignment and they did not learn from mistakes in people. “Only because they once made a mistake does not mean that they will never make it again,” D. Diesbach added.
It is especially dangerous for a tool like Grock, which De Diesbach says, “There is a more pronounced bias, which is very combined with the ideals propagated among others by Elon Mask.”
The problem is that musk has defective and fundamentally incredible to the incredible tools directly into a global city square and marketed it as a way to verify information. Failure is becoming a feature, not a bug, with dangerous consequences for public speech.
X did not immediately respond to any request for comment.