In a new essay and interview with MIT Technology Review, Bill Gates has issued his starkest warning yet about artificial intelligence, stating that the technology has crossed critical thresholds in bioterrorism, cybersecurity, job displacement, and control, and calling for immediate societal and governmental action to mitigate risks.
Bill Gates, the philanthropist and former Microsoft CEO, is sounding a new alarm on artificial intelligence, arguing that the world has already passed multiple dangerous thresholds without adequate safeguards. In an essay published today and an accompanying interview with MIT Technology Review, Gates said he is 'stunned at the lack of concern and discussion outside of the industry' regarding AI's growing capabilities.
'We've crossed the threshold in terms of [AI's] bio-capabilities, cyber-capabilities, psychosocial capabilities, job-market-destruction capabilities, and even the lack of control,' Gates said. He described himself as a 'shrill voice' trying to wake up leaders in government, industry, and civil society.
Perhaps most chillingly, Gates highlighted the risk of AI-enabled bioterrorism. He warned that any model capable of designing novel molecules should be monitored, estimating that bioterrorism risk is 'about 50 times more scary, more likely than a natural pandemic risk.' This places AI-driven biological threats far above the pandemic risks that have dominated global health discussions in recent years.
Gates also proposed novel policy ideas to address the coming disruption. Among them are 'human-reserved jobs' — positions that societies agree should be preserved for people, potentially varying by nation — and taxes on robots and AI-generated tokens. The robot tax concept, which he has floated before, would set aside revenue from AI-driven automation that replaces human workers.
Despite the dire warnings, Gates expressed cautious optimism about AI's potential to revolutionize agriculture, health care, education, and bureaucracy. He argued that humanity will eventually reach a state of abundance, but only after a period of significant turbulence.
Gates' intervention comes as frontier AI models continue to advance rapidly, with companies like OpenAI, Google DeepMind, and Anthropic releasing increasingly powerful systems. Critics note that regulation has struggled to keep pace, with few binding international agreements on AI safety or biosecurity.
The interview, conducted at Gates Ventures in Kirkland, Washington, paints a picture of a tech titan deeply unsettled by the trajectory of an industry he helped create. Whether his call to action will catalyze change remains an open question.
Analysis
Why This Matters
- Bill Gates, as a co-founder of Microsoft and a major philanthropist with deep ties to global health and technology, carries significant weight in policy circles. His shift from techno-optimism to urgent alarm could influence government regulation and public opinion.
- The specific focus on bioterrorism risk — which Gates compares to 50 times the danger of natural pandemics — highlights a threat that has received relatively little public attention compared to job displacement or autonomous weapons.
- Gates' proposals for human-reserved jobs and robot taxes could shape future political debates about economic redistribution in the age of AI, even if they face significant implementation challenges.
Background
Gates has been involved in AI discussions for decades, but his public stance has evolved markedly. In the early 2010s, he was broadly optimistic about AI's potential. By the late 2010s, he began warning about risks, but his tone was measured. His 2023 book and public statements emphasized both promise and peril, but always with an underlying belief that problems could be managed.
Today's essay marks a significant escalation. Gates now argues that the window for proactive intervention has closed on several fronts — that we are already living with dangerous capabilities that lack guardrails. This reflects a broader shift among some AI leaders, including Geoffrey Hinton and Yoshua Bengio, who have publicly expressed regret about the pace of development.
The timing is notable: frontier AI models are becoming more capable, and governments worldwide are scrambling to draft regulation. The EU's AI Act, passed in 2024, was a major step, but its enforcement remains uncertain. The US has no comprehensive federal AI legislation.
Key Perspectives
Bill Gates: The technology has crossed dangerous thresholds in biosecurity, cybersecurity, psychological manipulation, and economic disruption. He calls for immediate monitoring of models that can design novel molecules, as well as societal mechanisms like human-reserved jobs and robot taxes.
AI industry leaders: Many companies, including OpenAI and Anthropic, have their own safety teams and have supported some regulation, but they also argue that slowing development could cede advantage to competitors (including China) and delay beneficial applications in medicine and climate.
Critics and skeptics: Some experts question whether Gates' warnings are overly alarmist. David Leslie, a researcher at the Alan Turing Institute, has cautioned that 'fear-based narratives can lead to overregulation that stifles innovation without addressing root causes.' Others note that Gates' proposed solutions — like robot taxes — are politically unworkable in most countries and could hamper economic growth.
What to Watch
- Whether governments respond with concrete regulation on AI biosecurity, particularly around DNA synthesis screening and model evaluation requirements.
- The reaction from major AI labs: will they voluntarily restrict capabilities, or push back against calls for slower development?
- Upcoming international AI safety summits, such as the planned 2026 meeting in South Korea, where Gates' memo could influence the agenda.
- Public and political appetite for novel policies like human-reserved jobs or robot taxes, which would require significant legislative action.