Bill Gates has broken from the tech establishment's cautious hedging on AI risk, publishing a nearly 6,000-word essay that pairs domestic labor-protection mechanisms — a robot tax, a "human reserved" job category, and a token tax — with a direct bid for US-China AI safety diplomacy, including a planned meeting with Xi Jinping1,4,8. ANALYSIS The combination is unprecedented among major tech figures: no one of Gates's stature has publicly welded a concrete labor-market framework to a geopolitical negotiation agenda in a single policy document.
Why it matters
Gates is not merely sounding an alarm. He is proposing specific fiscal and regulatory instruments at a moment when, by his own account, "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". He told MIT Technology Review he is "just stunned at the lack of concern and discussion outside of the industry". ◆ That framing — thresholds already crossed, not approaching — sets a more urgent baseline than the incremental "responsible AI" rhetoric that has dominated industry discourse.
The big picture
The essay, titled "The turbulent AI era is here. The choices we make now are critical," argues that the world "is not preparing" for what is coming2,6. Gates's domestic proposals center on three interlocking ideas. First, a robot tax that corrects a tax-code asymmetry: "Right now, if you're an employer and you hire someone, you pay payroll taxes on their earnings. But if you buy a robot, you can usually write it off right away as a business expense. The tax system nudges you toward replacing people with machines," Gates wrote. Second, he envisions taxing revenue from companies using tokens or AI outputs to supplant labor. Third, he proposes designating certain jobs as "human reserved," likening the concept to nature reserves that protect the environment3. Gates said society should reserve up to 40% of current jobs for humans, while acknowledging "it is hard to get above 30% or 40% of jobs replaced by AI".
ANALYSIS The "human reserved" framing is notable because it shifts the policy conversation from retraining displaced workers after the fact to preemptively ring-fencing categories of employment — a structural intervention rather than a remedial one.
On the geopolitical strand, Gates told Reuters he believes China might agree to restricting potentially dangerous AI model releases if the United States took initiative first. "The whole world would go along with that," he said9. He also wants countries to monitor AI's ability to create molecules and carry out biological attacks, arguing that such monitoring "would not get in the way of any tech industry's development". His staff are working to finalize a meeting with Xi for his next China trip, tentatively scheduled for later this year.
Between the lines
Gates offered a pointed indictment of his own industry's candor. He told the New York Times that tech executives are privately "very worried" about AI disruption but publicly downplay the risks to protect fundraising and planned IPOs5. ANALYSIS That claim reframes the gap between private concern and public messaging not as mere caution but as a market-driven information asymmetry — one that Gates, no longer running a public company or raising venture capital, is uniquely positioned to expose.
He also argued that "you cannot count on an industry to self-regulate" and that the US government "is not a cutting-edge buyer or big R&D funder" in AI, meaning traditional regulatory leverage is weak. ◆ The implication is that novel fiscal tools — the robot tax, the token tax — are necessary precisely because the government lacks the procurement and funding power it wielded in earlier technology transitions.
Gates endorsed "Pacing The Frontier," the open letter from AI employees pushing for a slowdown, while expressing skepticism that a slowdown would be sustainable. ◆ That tension — endorsing the goal while doubting its durability — explains why his proposals lean toward economic incentives and international agreements rather than outright moratoria.
The Gates Foundation's own work illustrates the other side of his ledger: it is working with OpenAI to help healthcare workers in Rwanda and with Anthropic on AI-powered health checks for pregnancy. The foundation is aiming to spend $200 billion by 2045. ◆ Gates is thus positioning himself as neither accelerationist nor decelerationist but as a broker arguing that the benefits require guardrails to survive contact with reality.
What's next
The planned meeting with Xi, if finalized, would test whether Gates's bilateral framing gains traction at the state level. On the domestic front, the "human reserved" concept and robot tax are now in the policy conversation; their viability depends on whether legislators treat them as actionable proposals or rhetorical gestures. Gates said the safety net "will need more resources" and that the token tax is "key to funding" it. ◆ The next marker to watch is whether any of these instruments — job reservation, robot taxation, token taxation — appear in draft legislation or executive guidance, converting Gates's essay from provocation into policy scaffolding.