Microsoft ($MSFT) co-founder Bill Gates said he does not want the U.S. to slow AI innovation, but he believes stronger safeguards and monitoring are necessary as the technology becomes more powerful. Speaking on NBC’s “Meet the Press,” Gates dismissed the idea that a simple AI “kill switch” is enough to manage future risks. He said current systems are not yet autonomously taking over computers in ways humans cannot stop. Instead, Gates called for laws requiring companies to closely monitor advanced AI models and maintain detailed records of what those systems are doing. His concern is less about AI independently escaping control today and more about people using increasingly capable models for cyberattacks, bioterrorism or other dangerous purposes. Gates warned that AI could eventually be powerful enough to contribute to events causing catastrophic loss of life, saying the potential consequences could reach as high as “a billion deaths.” Asked what he would tell President Donald Trump, Gates said the U.S. should continue innovating quickly while establishing stronger oversight standards rather than imposing a broad slowdown. Trump has similarly argued against regulations that could significantly slow American AI development, saying the technology could become larger economically than the Industrial Revolution while maintaining that existing government authority can address serious misuse. Gates’ position therefore falls between the two extremes: keep advancing AI, but require stronger monitoring, accountability and safeguards before increasingly powerful systems become harder to manage.