Nadella: Govern AI models like insider threats. Microsoft CEO Satya Nadella urged companies to treat advanced AI models as insider threats, outlining seven principles including model diversity, containment, independent controls, and incident disclosure. In an essay on X, he argued firms must restrict AI authority, monitor actions independently, and retain human override capability, rather than trusting AI developers alone.