Bill Gates Says AI Self-Regulation Isn’t Enough and Governments Must Monitor AI
Microsoft co-founder Bill Gates has added his voice to the growing debate over artificial intelligence, saying the industry cannot be left to police itself. In an interview with NBC News’ “Meet the Press,” Gates argued that governments must be involved in setting and enforcing safeguards for advanced AI systems.
“No One Thinks Self-Regulation Is Enough”
Asked by moderator Kristen Welker whether self-regulation was sufficient or whether legislation was needed in Washington, Gates answered that no one thinks self-regulation is enough. When pressed on whether lawmakers should pass legislation, he said “absolutely.” He added that law enforcement and politicians need to join the discussion about what safeguards and monitoring should look like.
Monitoring Over Kill Switches
A central point of Gates’ argument is that a simple emergency shutdown is not an adequate answer. He said a kill switch has a role but cannot replace monitoring and safeguards built into AI systems. He described monitoring as watching what the AI is doing, and said it should be introduced in every AI lab. Gates also said advanced models need oversight before a dangerous incident occurs rather than relying only on an emergency mechanism afterward.
A Stark Warning About Risk
Gates did not downplay the stakes. He said AI is certainly powerful enough to drive events that could cause a billion deaths. He also said there has never been a weapon as powerful as the combination of people with ill intent and the latest AI tools. He pointed to dangerous thresholds he said were crossed this year, including capabilities linked to biological and cyber attacks.
A Knowledge Gap in Government
Gates noted that governments usually understand dangerous technologies because they funded the research or are major customers, as with nuclear weapons or rockets. AI, he said, grew out of the private sector, leaving governments with little expertise. He criticised the framing of AI as a race between nations, saying it leads some people to demand full speed ahead, and called it completely irresponsible not to require safeguards and monitoring in every AI system. He also said governments requiring such measures would not slow the industry’s progress.
Global Cooperation Will Be Hard
Gates warned that getting countries to agree on AI rules will be harder than the Cold War-era nuclear negotiations. He argued that unless governments insist on safeguards, the world will see a lot of big negative outcomes. His remarks came as he addressed President Donald Trump’s stance against regulating AI.
A Wider Push for Oversight
Gates is not alone. Sam Altman of OpenAI told a United Nations gathering that international standards are needed for measuring capabilities, assessing risks and preserving meaningful human oversight. Microsoft President Brad Smith said AI companies do not need to wait for federal law before slowing themselves down. Anthropic CEO Dario Amodei has also called for frontier developers to slow the pace, though only to a point where they still lead China, and for global coordination.
At the state level, governors including Wes Moore and Kathy Hochul have released plans to boost AI regulation, and Hochul’s plan creates a new office to oversee compliance.
Congress Yet to Act
It remains unclear what Washington will do. Republican congressional leaders have so far shown little urgency on AI legislation, and critics of regulation argue that companies should be allowed to govern themselves.
What Comes Next
With business leaders, state officials and now Gates calling for oversight, pressure on federal lawmakers is growing. The open question is whether governments can build the expertise and the international agreement needed before the technology moves further ahead.
About the Author
