Gates calls for AI safeguards, warns global agreement will be difficult

Bill Gates called for legislation, monitoring and other safeguards around artificial intelligence, warning that international agreement on AI rules could be harder to achieve than Cold War-era nuclear arms negotiations.

In an interview with NBC News’ “Meet the Press,” the Microsoft co-founder said it would be “completely irresponsible” not to require AI systems to include safeguards and monitoring capabilities. He identified malicious use of AI as the most immediate danger, pointing to potential bioterrorism and large-scale financial fraud.

Gates said he would not oppose a “kill switch” for advanced AI systems, but argued that such a measure would not fully address the threat posed by people using the technology with harmful intent. He supported involving lawmakers and law enforcement in determining what protections AI companies should adopt, saying regulation could add overhead without dramatically slowing the industry.

His comments came as lawmakers considered new federal controls. Democratic Rep. Ro Khanna said he would introduce the “Human Control Over AI Act,” which would prohibit recursively self-improving systems and autonomous changes to core objectives or shutdown controls until federal safeguards and approvals are established. According to CNBC, the proposal would also create a federal AI agency, licensing and audit requirements, security standards and penalties for disabling safeguards or deploying unauthorized systems.