THE CONVENTIONAL WISDOM among Fortune 500 CEOs is that artificial intelligence will enhance workers, not replace them — that the correct posture is optimism tempered by vague references to "reskilling." Jamie Dimon's 2025 shareholder letter, published Monday, breaks from this script with unusual candor. The head of the world's largest bank by market capitalization — a firm spending $19.8 billion on technology this year and explicitly trying to become "fundamentally rewired" for the AI era — concedes that AI deployment "may move faster than workforce adaptation to new job creation." Then he proposes a specific policy response: double the Earned Income Tax Credit and convert it into a monthly negative income tax. Coming from a CEO whose bank posted $57 billion in net income last year, it is a signal worth parsing.
The letter is, as always, expansive — nearly 60 pages covering geopolitics, bank regulation, trade policy, and a Europe section that reads like a foreign-policy white paper. But threaded through the sprawl is a remarkably detailed AI thesis. Dimon calls AI genuinely transformational and argues its adoption will likely be far faster than prior technological shifts like electricity or the internet, which took decades to roll out. He forecasts hyperscaler AI capital spending jumping from $450 billion in 2025 to roughly $725 billion in 2026 — a 61% increase in a single year — and says the investment is "not a speculative bubble." He acknowledges AI will "definitely eliminate some jobs" while enhancing others, and states that JPMorgan has "definitive plans" for supporting and redeploying affected employees.
The $20 billion confession
Yet the letter's most striking passage is the one where Dimon's dual roles — as AI accelerator and AI worrier — collide. He writes that in prior technological transformations, "labor had time to adjust and retrain," but concedes that AI may not afford the same luxury. This is not an abstract concern for JPMorgan. At an investor meeting in February, Dimon revealed that the bank has already displaced workers due to AI and has launched what he called "huge redeployment plans" to move them into other roles. The bank's overall headcount held steady at roughly 318,500 over the past year, but the composition shifted beneath the surface: operations staff fell 4%, support roles dropped 2%, while client-facing and revenue-generating positions grew 4%. Operations teams now handle 6% more accounts per employee, fraud-related costs per unit have dropped 11%, and software engineer productivity has risen 10%.
JPMorgan, in other words, is already living the transition it warns about — and it has the scale and the $20 billion tech budget to absorb the friction internally. Most employers do not. A Deloitte survey of over 3,200 senior leaders found that while 53% are focused on raising AI fluency, only 30% are actually redesigning their organizations around AI-driven workflows — the structural work that JPMorgan is actively doing. The gap between "educating the workforce" and "redeploying the workforce" is precisely where Dimon reckons the social risk accumulates.
Which is why the policy prescriptions in the letter are so pointed. Dimon proposes roughly doubling the EITC and eliminating the child requirement, converting it from an annual tax credit to an effective monthly negative income tax — a mechanism that would put money directly into the pockets of lower-income workers on an ongoing basis rather than in a lump sum at year's end. He frames it alongside calls for federal outcome reporting for schools, expanded skills training, and certification programs in fields from cybersecurity to advanced manufacturing. He notes that approximately 23% of American workers earn less than $17 an hour and that close to 10% have incomes of $20,000 or less. These are the workers least equipped to navigate a rapid AI-driven labor transition — and the ones most likely to fall through the gap between displaced roles and whatever emerges on the other side.
The broader context makes Dimon's position more significant than it might otherwise appear. BlackRock CEO Larry Fink warned in his own shareholder letter of the "real risk" that AI widens wealth inequality. A group of economists and investors has begun pitching Washington on an AI displacement "circuit breaker" — automatic stabilizers that would trigger expanded income support if AI-related job loss spikes beyond certain thresholds. A recent Brookings analysis identified 6.1 million U.S. workers — concentrated in clerical and administrative roles, 86% of them women — who face both high AI exposure and low adaptive capacity. The World Economic Forum projects 92 million jobs displaced globally by 2030, offset by 170 million new roles, but the net-positive headline masks a severe distributional mismatch: the jobs being destroyed and the jobs being created require different skills, pay different wages, and exist in different geographies.
Dimon made the point viscerally at the February investor meeting with a thought experiment. What if autonomous vehicles eliminated 2 million commercial truck drivers overnight — drivers earning an average of $120,000 a year? The replacement jobs, he suggested, might pay $25,000 stocking shelves. The question he posed was not whether it was technically possible but whether society would accept the human cost. The fact that this question is being asked publicly by the CEO of a bank that just delivered record revenue for the eighth consecutive year — a bank aggressively deploying the very technology in question — suggests the conversation has moved past the theoretical. Whether Washington is listening is another matter; the EITC has not been meaningfully expanded since 2009. Dimon seems to be betting that the AI cycle will force the issue before the political one resolves it. History suggests he is probably right about the timing — and probably optimistic about the response. ■
For more, join 75,000 subscribers getting tech's favorite brief
