• Bill Gates warns AI is already powerful enough to enable events causing a billion deaths, emphasizing the combination of advanced tools and malicious actors.
  • He calls for mandatory government oversight, arguing that self-regulation is insufficient and that safeguards can be implemented without halting progress.
  • The warning comes as Microsoft (MSFT) reaps record gains from AI, highlighting a growing tension between commercial success and catastrophic risk.

Gates: 'Certainly Powerful Enough'

Bill Gates warned that artificial intelligence is already capable of enabling catastrophes that could kill a billion people, according to remarks released ahead of a Meet the Press interview. The Microsoft co-founder said the technology is “certainly powerful enough” to drive such events, pointing to the combination of increasingly advanced tools and individuals with malicious intent.

No weapon in history has matched that combination’s potential scale, Gates argued. He described the risk as a scenario rather than an imminent prediction, but insisted the current trajectory carries significant dangers. His central concern is that capable AI could empower cyberattacks, engineered biological threats, fraud, disinformation, and attacks on critical systems unless governments impose enforceable safeguards.

'No One Thinks Self-Regulation Is Enough'

Gates called for U.S. legislation and direct governmental involvement in setting mandatory monitoring and safety rules. “No one thinks self-regulation is enough,” he said, characterizing compliance as manageable operational overhead rather than a requirement to halt AI progress.

The appeal lands amid a fragmented global regulatory landscape. In the United States, congressional proposals on AI have proliferated but major federal legislation has largely stalled amid partisan disagreements. The European Union’s AI Act is already moving from rulemaking to enforcement, with transparency requirements taking effect in August 2026 and providers of general-purpose models facing obligations around risk assessment, incident reporting, and cybersecurity. The United Nations has established an advisory scientific panel and a global dialogue on AI governance, though its role remains non-binding.

Gates’s push for government intervention marks a shift in his recent public stance. In August, he warned that governments and the technology industry were underestimating AI-related risks to employment, cybersecurity, biosecurity, social cohesion, and human relationships.

Microsoft's AI Boom vs. Gates's Warning

The headline concerns Gates personally rather than a current Microsoft corporate announcement, but it underscores a striking tension. Microsoft, led by Chairman and CEO Satya Nadella, has become one of the biggest beneficiaries of the AI boom. For fiscal 2026, the company reported $331.8 billion in revenue, up 18% year over year, and $155.2 billion in operating income, up 21%. Microsoft Cloud generated $214 billion, up 27%, while Azure surpassed $100 billion in annual revenue and grew 41%. Microsoft 365 Copilot exceeded 30 million paid seats.

That commercial success illustrates the broader policy challenge: governments are being asked to regulate a technology that is simultaneously a major source of investment, productivity hopes, and corporate revenue. AI has become a key driver of cloud spending, corporate capital expenditure, and strategic competition.

Gates has not abandoned AI’s potential benefits. On September 14, the Gates Foundation committed at least $1 billion over two years to expand beneficial AI access, particularly in education, health care, agriculture, and language-data infrastructure for underserved communities.

Uneven Gains and a Widening Divide

Gates has warned that economic gains from AI will not necessarily be evenly shared. He has said the technology may displace jobs across white- and blue-collar sectors, put particular pressure on entry-level employment, and create social and political strain if policy fails to prepare workers and safety nets. In earlier 2026 proposals, he advocated an AI “token tax” to reduce incentives for indiscriminate labor replacement and fund support for affected workers, as well as “Human Reserved” roles where human involvement would be required. Those are Gates’s proposals, not enacted policy.

For low-income countries and underserved groups, the risk is a widening digital divide: models often work less well in local languages and settings with sparse data. The Gates Foundation’s $1 billion plan aims to counter that gap through tools for teachers, frontline health workers, and smallholder farmers.

The near-term market implication is likely continued investment in AI compute, cloud capacity, models, and applications — alongside rising costs for evaluation, cybersecurity, incident reporting, model documentation, and compliance. Longer term, firms that can demonstrate safety, traceability, resilience, and responsible deployment may gain an advantage as regulation becomes more enforceable.

Public debate is likely to center on whether Gates’s worst-case framing is proportionate, which technical safeguards actually work, who should bear compliance costs, and whether oversight should focus primarily on frontier model developers, downstream deployers, or both. Critics of broad AI restrictions often warn that overly rigid rules can entrench the largest incumbents or push research to less-regulated jurisdictions; advocates counter that voluntary commitments are insufficient for risks with potentially global consequences.

Microsoft did not respond to a request for comment on Gates’s remarks. Gates’s office could not be immediately reached.

A previous version of this article misstated the timing of the EU AI Act’s transparency requirements. They took effect in August 2026.