In his "Meet the Press" interview that aired over the weekend, the billionaire told Kristen Welker that malicious actors could use AI to cause unprecedented chaos and death by enabling biological or chemical weapons, accelerating cyberattacks, increasing the risk of nuclear conflict, empowering autonomous weapons, or eventually becoming too difficult for humans to control.
"AI is certainly powerful enough to drive events that cause a billion deaths," he said. "There's never been a weapon as powerful as the combination of people with ill intent using the latest AI tools."
In a year or so, Gates said malicious actors might not even be necessary.
"Not only do we have to stop bad people from doing things, but we have to stop AI itself from getting out of control," he warned. "Can an AI find flaws in code to attack and shut down electricity or hospitals or to scramble banking systems this year?"
Believing that is not too far off, he has joined the growing number of tech experts who want the government to step in and slow things down.
In the July 2026 "Pacing the Frontier" statement, more than 1,200 employees at major AI companies call for governments to develop mechanisms that could slow the pace of AI development if it becomes too rapid. Notable signatories included Anthropic CEO Dario Amodei, OpenAI chief scientist Jakub Pachocki, Meta chief scientist Shengjia Zhao and Google DeepMind AI-safety leader Anca Dragan.
"No one thinks self-regulation is enough," Gates said. "You need law enforcement and the politicians to get into the discussion about what safeguards and monitoring look like."
He said humans still have the power to stop AI from its worst impulses, but the development of Artificial General Intelligence, a theoretical AI system that matches or exceeds human cognitive abilities across any intellectual task, is on the way. When it is achieved, it might be too late to put the genie back in the bottle.
Pointing out that humans have created an intelligent species that will soon be "insanely smarter than we are," Gates reiterated that "it's completely irresponsible not to require every AI to have these safeguards and monitoring capabilities."
