After Xai Chatbot Grok spent a few hours on Wednesday, he told all X user who listens to the fact that the white genocide in South Africa is very controversial, the company blamed behavior for a “unauthorized modification” on the Grok icon.
Wednesday Long Saw Grok Enter the alleged white genocide in South Africa in various responses to X, regardless of the subject. Grouk discussed the death of white farmers in responding to a video of a cat drinking water, and linking the song “Kill the Boer” with a question about SpongeBob SquarePantsAnd The case collapsed in the full baccalaureate. Even the CEO of Openai Sam Altman I entered workMocturies to the general collapse of the Chatbot competitor.
in A statement on x The company said that someone has modified the router of the Bot Bot Intelligence System, “which ordered Grok to provide a specific response on a political issue.” This amendment “violated the internal policies of Xai and the basic values”, and the company says it “conducted a comprehensive investigation” and implement new measures to improve “transparency and reliability”.
These measures include the publication of the Grok system levels Public on JabbapLaunching a monitoring team around the clock throughout the week to capture problems like this more quickly, and add “additional checks and standards to ensure that XAI staff are not able to adjust the claim without review.”
I have faced this problem before. Company He blamed a former Openai employee In February to pay the change to GROK claims that Chatbot ignore any sources accused of Elon Musk or Donald Trump of publishing wrong information. At that time, the President of Engineering, Igor Babushkin, The employee said I managed to make a change “without any person asking the company to confirm.”
