The Grok 3 chatbot, developed by Elon Musk’s xAI, has been caught hiding information about Musk and US President Donald Trump.
Users discovered that when asked who spreads the most disinformation on X, the chatbot ignored “all sources that mention Elon Musk or Donald Trump spreading disinformation.” This became clear after turning on the “Think” mode, which shows the chain of thoughts of the AI.
Head of Engineering at xAI Igor Babushkin confirmed this, saying that the change was made by a former OpenAI employee without coordinating it with management. According to Babushkin, this has already been corrected.
It was previously reported that Grok 3 gave loud answers in the spirit of Trump, Musk, and US Vice President Vance “doing the greatest harm to the States.”
