Microsoft co-founder Bill Gates warned that artificial intelligence is powerful enough to drive events causing a billion deaths if misused. Speaking in a September 2026 interview with NBC News, Gates argued that self-regulation by technology companies is insufficient and called for mandatory government and law enforcement oversight.
The debate over artificial intelligence shifted following stark warnings from prominent figures in the tech sector. While software developers and executives publicly debate safety guardrails, a wave of resignations among AI researchers has brought internal fears into public view. Former and current employees at leading firms argue that current commercial incentives encourage companies to downplay the catastrophic risks associated with advanced autonomous systems.
Bill Gates Warns of Billion-Death Risk From Misused AI Systems
In a broadcast clip released by NBC News from his “Meet the Press” interview, Gates stated plainly that the technology carries destructive potential on a massive scale. When asked about predictions that autonomous software could wipe out humanity, the Microsoft co-founder explained that while eradicating every human being remains difficult, malicious actors armed with cutting-edge tools could orchestrate unprecedented devastation.
Bill Gates said that AI is powerful enough to drive events that could cause a billion deaths, according to Expressen.
AI is definitely powerful enough to drive events that could lead to a billion deaths, Gates told NBC News, noting that history has produced no weapon comparable to the combination of malicious human intent and advanced machine learning models.
Why Tech Executives and Researchers Are Sounding Alarms Behind Closed Doors
Gates’ comments align with a growing internal rebellion across the artificial intelligence industry. Major departures have punctuated the sector as researchers clash with leadership over safety priorities. Jacob Coxon left Anthropic after concluding that the firm and its competitors are gambling with our lives. Evan Hubinger stated that they really do earnestly believe AI could kill all humans, added in a post on X.
Similar departures mark other industry giants. In 2024, Jan Leike and Daniel Kokotajlo quit OpenAI over safety concerns. This year, Anthropic safety chief Mrinank Sharma departed, warning the world is in peril
. Alex Turner left Google DeepMind after it signed a deal with the Pentagon permitting “killer drones”. These exits reflect a broad consensus among technical staff that corporate self-policing fails to address the speed and autonomy of modern neural networks.
The Limits of Corporate Self-Regulation and the Call for State Oversight
Addressing the ongoing debate over voluntary codes of conduct, Gates dismissed the idea that tech companies can effectively monitor themselves. Behind closed doors, he noted, developers are aware of the risks, and tech companies have intentionally toned down the risks to make money. In an interview with The New York Times, Gates said that behind closed doors, those who understand how good this technology is and how fast it is developing are very worried.
Bill Gates said that nobody thinks self-regulation is sufficient, according to Expressen.
To counter these risks, Gates argued that regulatory mandates involving law enforcement agencies and elected officials must become mandatory. They need to participate in the discussion of what safeguards and monitoring look like, and it has to be a requirement. While acknowledging that external oversight will impose administrative burdens on software developers, he insisted that it will not slow down what they are working on.
Balancing Global Development Goals Against Catastrophic Risks
The existential warnings arrive alongside separate philanthropic announcements regarding the technology’s immediate utility.
He added that AI is different, representing a technology moving faster and reaching more people than anything before it, and because of that, no one knows exactly what the future will look like even five or ten years from now, including him. AI is already revolutionising global development in all sorts of ways, with medical imaging and image reading techniques making diagnostics more precise.
Where Industry Consensus Stands on Extinction Risks
The debate over catastrophic harm is no longer confined to academic speculation or science fiction forums like LessWrong. Executive leadership across competing firms previously acknowledged the scale of the threat.
Whether regulatory bodies can establish enforceable controls before autonomous capabilities outpace human oversight remains the central unresolved question facing policymakers, as businesses worldwide scale back AI tools, citing ‘high costs’ and ‘low returns’.