Bill Gates Says AI Could Be the Greatest Equalizer or the Worst Source of Injustice
In a nearly 6,000-word essay, the Microsoft co-founder warns the world is unprepared for what AI will do to jobs and governments, and calls for a new international body to manage the fallout.

Key points
- Bill Gates published a nearly 6,000-word essay warning that the shift to an AI-driven world will be "one of the most turbulent times in human history."
- Gates predicts many jobs will disappear permanently, affecting office workers and factory and delivery workers alike.
- He proposes taxing AI tokens (the units of text an AI system processes) and robots to slow job replacement and fund social safety nets.
- Gates calls for a new international organisation to govern AI risks that cross national borders, and says it would require US-China cooperation.
- He vows to personally lobby lawmakers in Washington and with world leaders to put AI and inequality higher on the political agenda.
Bill Gates is done staying quiet.
The Microsoft co-founder published a nearly 6,000-word essay this week that reads more like a warning siren than a technology preview. Covered by The Verge AI, the piece is titled "The turbulent AI era is here. The choices we make now are critical," and its central message is blunt: the world isn't ready, and it's not preparing fast enough.
"AI will either be the greatest equalizer ever invented, or the worst source of injustice," Gates writes. That's a sharp shift from the hopeful tone of his 2023 essay on the same subject. We covered his argument that AI has already crossed its danger thresholds earlier this year.
What does Gates actually propose?
He offers two concrete policy ideas, notable because concrete proposals are rare in essays of this kind.
First, Gates wants governments to tax AI tokens (the units of data an AI system processes each time it answers a question) and physical robots. If using AI or a robot carries a cost, companies will think harder before replacing a human worker. Revenue would go into social funds supporting people who lose their jobs.
Second, he proposes a category he calls "Human Reserved": specific jobs governments would protect from AI replacement entirely. He doesn't list which jobs qualify, saying he'll share details in future writing. The intent is to set a floor the technology cannot cross.
He also wants a new international organisation to handle AI risks that spill across borders. Gates compares it loosely to agreements governing nuclear weapons, commercial aviation and ozone-layer protection. Four such precedents, he argues, show it can be done. He acknowledges this would need the United States and China to cooperate, but offers no mechanism for achieving that.
Should ordinary workers be worried?
Gates thinks the threat to employment is real and lasting. "Many jobs will disappear forever," he writes, covering white-collar roles and blue-collar ones. He urges political leaders to act before unemployment rises sharply, not after.
If your work is largely repetitive or involves processing standard information, this essay suggests your industry is already in the crosshairs. Gates offers no timeline, though he's clear the disruption will arrive faster than most people expect. And he's candid about his own view: "if someone had a credible plan for slowing down AI advances globally, I would likely support it."
What happens next?
Gates promises follow-up writing on how his foundation is using AI and on specific ideas for ensuring benefits outweigh harms. He commits to raising the issue with lawmakers every time he visits Washington or meets world leaders.
The essay signals that one of tech's most recognisable figures intends to be loud in AI policy debates. Whether it translates into change depends on whether governments move, as Gates puts it, before "communities are hurting and public trust has eroded." My read: the proposals are serious in intent but thin on mechanism, and the international-body idea in particular will go nowhere without a concrete answer to the US-China question Gates himself raises.
Common questions
What is an AI token?
A token is the basic unit an AI system uses to read and write text, roughly equivalent to a syllable or short word. Taxing tokens means charging companies a small fee each time their AI processes language, making mass automation slightly more expensive.
Which jobs would "Human Reserved" protect?
Gates doesn't specify here. He says he'll share more details in future writing, but the concept is a legal or regulatory category that would bar AI from fully replacing humans in certain roles.
Is this a law yet?
None of it is. These are proposals; no government has committed to any of these measures.



