Anthropic CEO Dario Amodei apparently can’t believe that not everyone at the company is as noble-minded as he is. Now, seemingly contrary to this “safety-first” mission, Anthropic has been moving to take itself public in an anticipated trillion-dollar IPO. Earlier this year, it dropped a critical safety pledge to stop training an AI model if it couldn’t guarantee it had proper guardrails in place. And despite making a show of clashing with the Pentagon over concerns about the safe deployment of its AI systems, Anthropic also let the military use its Claude AI to select targets in Iran. More on Anthropic: A Whole Bunch of People’s Claude Chats Are Publicly Accessible Online, and There’s Some Wildly Private Stuff in There