Bill Gates told Semafor in an interview and reiterated in a blog post published on Wednesday that it is time to sound the alarm about artificial intelligence. He said AI has crossed dangerous thresholds in biology, cybersecurity, white‑collar work and human relationships.
This marks a notable shift from Gates’s position three years ago, when he predicted a "bumpy" but "manageable" job transition and argued that AI would make workers more efficient. He now warns that the speed of AI advances combined with what he describes as a near‑absent response from governments and tech leaders could produce economic catastrophe and "far fewer" jobs than exist today.
"I am in a state of shock that I’m sort of the first one saying, 'This is crazy. This is insane,'" Gates said. "I’m just deafened by the silence."
Why Gates is pressing for policy action
In his blog post and the interview, Gates urged technology and political leaders to act more urgently against AI threats. He warned that unchecked development could provoke a social backlash that would cripple AI’s potential benefits—especially in science and health—if people respond to mass job losses by turning against the technology. "What is the backlash going to be as these jobs go away?" he asked. "It’ll make the data center thing look like nothing."
Gates said this is the first in a series of planned essays; future pieces will include a deeper examination of biological risks posed by AI. "I wrote a whole book about how to avoid the next pandemic," he added, and argued that bioterrorism risk is now "50 times the natural pandemic risk."
He also said some technologists privately share his concerns, naming Microsoft AI CEO Mustafa Suleyman as someone who feels similarly. Demis Hassabis, co‑founder of Google DeepMind, has also stepped up warnings and proposed a federally overseen frontier‑AI standards body, funded mainly by industry, empowered to test advanced models and potentially coordinate slowdowns among leading labs.
Evidence and debate over the scale of job impacts
Academic and policy research shows mixed results about net job losses but supports Gates’s point about the breadth of AI’s impact. The International Monetary Fund (IMF) estimates AI could affect nearly 40% of jobs globally and about 60% in advanced economies. Stanford researchers found no economy‑wide displacement but reported a 19% employment gap for 22‑ to 25‑year‑olds in exposed occupations, largely due to reduced hiring. Yale’s Budget Lab has not yet identified a clear AI‑related labor‑market footprint.
Gates acknowledged his remarks are a forecast rather than a description of today’s headline employment figures. He said his worry is not primarily speed but scope: "There isn’t a job that isn’t affected," he predicted, adding that in most occupations humans will ultimately be "vastly inferior to the AI doing the job." Even if adoption is gradual, the widespread nature of the change would produce enormous aggregate effects.
Neither philanthropy nor industry can solve this alone, he says
Gates argued that governments must play a central role in responding to AI disruption. He dismissed the notion that philanthropy—despite the Gates Foundation’s roughly $10 billion a year in spending—can substitute for public policy: "Government is 20 times bigger than all philanthropy, so no, philanthropy cannot solve the safety‑net issue."
He warned that private companies are ill‑equipped to manage societal risk: model builders are "not policy people," he said, and they have "a little bit of a bias against talking about the potential negatives of their products." "The response can’t be, ‘Hey industry, you solve this,’" Gates said.
Policy proposals Gates outlines
Gates offered several concrete ideas for policymakers:
-
Designate certain occupations or services as "human‑reserved," meaning they would remain the domain of people even if machines could perform them more efficiently. Gates argued that market signals should ultimately serve humans and that declaring some tasks human‑reserved could be appropriate.
-
Impose a usage tax on AI compute (tokens) and robots. Gates said a use‑based tax would change incentives for firms to replace workers more directly than a corporate tax, which only captures profits and may be limited in fiercely competitive markets. He noted payroll taxes already apply to labor, while robots can be depreciated and avoid Social Security taxes. "Today, we are very pro‑robot in our tax policies, and I’m saying I think that’s a mistake," he said.
Gates also cautioned that the companies building the most consequential AI systems might not capture the majority of the economic gains. Heavy competition could commoditize models so that technological change reshapes the world without producing obvious profit winners. At current commodity levels, users may pay for execution but little for a model’s intellectual property, which could shift economic benefits to customers and downstream businesses instead.
He rejected the idea that his proposals are motivated by financial self‑interest and argued against the industry’s "accelerate at all costs" mentality. "The goal of humanity is not economic optimization," he said.
Where the debate stands
Gates’s warnings have found allies among other leaders. In 2023 Gates, Mustafa Suleyman, Demis Hassabis and hundreds of other experts signed a statement urging that the extinction risk from AI be treated as a global priority alongside pandemics and nuclear war. The core question remains whether societies and governments can design regulation and social protections to mitigate AI’s rapid diffusion and its potentially severe social consequences.
Gates said he will publish more essays on these topics, including a detailed analysis of AI‑related biological risks, and he called on both technology and political leaders to respond with greater urgency.
