In a series of X posts, AI Chatbot Grog apologies for what is claimed to be “terrible behavior.”
Posts appear to be an official statement from Xai, the Elon Musk Company Company in the back of the groke, as opposed to a GROC-explained. (Xai just got xwhere the GRKONG is displayed.)
Grok’s latest controversy comes after stated in the musk he wants to do the chatbot less “political in short order, Chatbot makes posts criticized Democrats and Hollywood executors, “ Repeated antisemitic memesand even Expressing Support for Adolf Hitler and referred to himself as “mechahitler.”
As a result, Xai removes some groke posts, temporarily Took Chatbot Offlineand updated the system of public prompting it.
Turkey is also Chatbot is prohibited to insult the president of the country, and x CEO Linda Yaccarino again announces that he is Go through this weekAlthough his notice does not mention the latest controversy of the groke and his departure reported months to make.
So after all, on Saturday, xai said“First, we apologize for the terrible behavior experienced by many.” The company blamed an “update of a passage code in @grok bot,” which emphasized “independent of the language language model of the @grok language model.”
This update made with groke “easily with X user posts; including such posts containing extremist views.”
Added to Xai That an “unexpected action” brings to withdraw the instructions such as, “tell you it and you are not afraid to offend people right.”
Company explanations explains Musk comments early this week Claiming that the groke “also follows the user’s prompts” and “also want to be pleased with and manipulate.”
Xai’s posts don’t talk about reporting techcrunch and other checking chains-of-minded summaries for reasonable grok 4, found that the latest version of Chatbot Seems to consult musk views and social media posts before responding to controversial subjects.
And the historian Angus Johnston pushes back to the idea that repeat posting the offensive content. she wrote to Bluesky that Xai and Musk’s explanations are “easy lies.”
“One of the most widely shared examples of grok antisemitism is initiated by the groke without a pre-born posting of thread – and with many users who pushed the groke,” Johnston said.
In recent months, there is also a groke posted repeatedly about “white genocide,” expresses doubt about the death of the Holocaustand In short censorship of unbearable facts about musk and his ally Donald Trump. In cases, XAI blames “unauthorized” changes and employees in rogue.
Despite the controversy, musk says The groke comes in Tesla vehicles Next week.
