Bill Gates has warned that artificial intelligence could enable bioterrorism on a scale that makes nuclear weapons “look like nothing,” calling it the most dangerous thing humans have ever approached.

In an interview with the New York Times’ Ezra Klein this week, Gates said AI has already “crossed the threshold” of being able to help a bioterrorist kill hundreds of millions of people. Speaking on NBC’s Meet the Press, he added that AI is “certainly powerful enough to drive events that cause a billion deaths.”

Gates argued that the Cold War deterrence model does not translate to AI-driven biological threats, since pathogen design increasingly runs on software rather than state-controlled materials. He said governments lack the technical expertise to manage the risk because AI capabilities emerged from the private sector.

He called for mandatory monitoring and record-keeping in AI systems, saying self-regulation and kill switches are both insufficient. Gates said he hopes to discuss the issue with US President Donald Trump.