Russian media reveals: US military uses AI in Iran airstrikes.

According to the Russian newspaper website on March 2, citing a report from the American Wall Street Journal, anonymous sources said that the US armed forces used the large language model "Claude" from the American company Anthropic during airstrikes on Iran, even though President Trump had ordered federal agencies to stop cooperating with the company on February 27. It is said that the Pentagon uses artificial intelligence to process intelligence, identify targets, and simulate various scenarios. This indicates that Anthropic's model has been integrated into the US military's decision-making system. The model was also reportedly used in a raid on the residence of Venezuelan President Nicols Maduro. A few days before the attack on Iran, Anthropic CEO Dario Amodei stated that the company had its own "red lines" regarding the use of AI. Information technology experts strongly oppose integrating AI into lethal weapons, but allow its use in other defense areas. On the eve of the escalation of the situation, the Israeli newspaper Jerusalem Post asked several popular large language models for the possible date of the attack on Iran. Claude predicted March 7th to 8th. The "Gemini" model did not respond immediately but later predicted March 4th to 6th. ChatGPT initially indicated a threat of attack on March 1st, with tension continuing for a week, and then narrowed the "window" to the night of March 3rd to the early hours of March 4th. The closest to the actual date was Elon Musk's chatbot "Grokk," which twice predicted February 28th.