Gates Calls for Mandatory AI Monitoring, Warns of Nation-State Attack Risk

Bill Gates has argued that AI has transferred the ability to cause serious harm, previously held by large nation-states, to small independent groups. Speaking on The Verge's Decoder podcast with Nilay Patel, Gates said this applies to both profit-motivated cyberattacks and destruction-motivated biological threats. He described the absence of mandatory model review or coordinated safeguard development as "stunning." Gates has called for powerful open-weight models to be restricted to government-certified platforms by law, where attempts to remove safeguards could be detected and blocked, while still allowing developers to use and customize capable models for free on those platforms. In a separate NBC interview covered by Politico, Gates said a kill switch requirement alone is insufficient to prevent AI-enabled attacks, and called for mandatory monitoring of powerful AI systems.