Jamie Dimon is not an AI sceptic. The JPMorgan Chase chief executive is deploying artificial intelligence across America’s largest bank and expects it to deliver a more productive society. But his warning about the technology — delivered at the World Economic Forum in Davos and recirculating widely this week — is about pace: roll AI out faster than society can absorb, he argues, and the result could be civil unrest.
Dimon’s position, as reported by Computer Weekly, is blunt about the destination. AI will eliminate some jobs, change many more and create some new ones. The open question is only how the transition is managed.
The truck-driver test
His illustration has become the most quoted part of the remarks. The United States has around two million commercial truck drivers. If autonomous technology made it possible to replace them all at once, drivers earning, in his hypothetical, $150,000 a year might find their next job pays $25,000. Do that to two million households overnight, Dimon argued, and “you will have civil unrest — so phase it in”.
The numbers are deliberately stark rather than statistical forecasts, and commentators have noted the warning is conditional: it is an argument about sequencing, not a prediction of collapse. But it is striking to hear the case for deliberate slowness made by the head of a bank that is itself investing billions in the technology.
What “phasing in” would actually mean
Dimon’s prescription is a joint programme between government and business, planned before displacement peaks rather than after: retraining for affected workers, relocation support, income assistance and early-retirement options — the same toolkit used when a town loses a factory. He has suggested local-level incentives for companies to slow deployments where necessary, and said JPMorgan would accept such arrangements if needed “to save society”.
He is equally candid about his own institution: JPMorgan, which employs more than 300,000 people, will likely have fewer employees in five years because of AI, even as it continues to grow globally. The difference he draws is between a workforce that shrinks through attrition and redeployment over years, and one cut abruptly for quarterly savings.
Why the argument is landing now
The remarks are resurfacing because the debate has moved from theory to headcount. Companies across sectors are citing AI in restructuring announcements, and the productivity story in markets — visible in everything from banks’ trading results to the chip-export boom in global trade data — is running well ahead of any agreed framework for the workers displaced along the way.
Dimon’s framing reframes AI adoption as a political-economy problem rather than a purely technical one. The technology’s benefits, on his account, are real but unevenly timed: productivity and profits arrive quickly, while new roles, retrained workers and adjusted communities take years. Managing that gap — through pacing, safety nets and honest planning — is, in his telling, the difference between a transition societies accept and one they resist.
Whether governments take up the offer of partnership is another matter. But when the most powerful banker on Wall Street argues that moving too fast could destabilise the society his industry depends on, the argument is difficult for boardrooms to dismiss as anti-technology sentiment.