Jamie Dimon's Forty Phone Calls

Cover photo: Jamie Dimon, CEO of JPMorgan Chase, by Steve Jurvetson, licensed under CC BY 2.0.

Jamie Dimon did not delegate this one. In July, according to Reuters, JPMorgan's CEO started personally phoning the chief executives of more than 40 companies, asking them to join a coalition that would sit down together and work out what AI actually does to a bank, a power grid, a rail network, a phone system, before something breaks that nobody planned for. Financial services, energy, water, telecoms, airlines, railroads. Industries that don't normally share a room, sharing one now, because the person calling them is the head of the largest bank in the country and he doesn't tend to make forty personal calls about things he considers optional.

The coalition already existed, technically. JPMorgan, Mastercard, AT&T, Berkshire Hathaway Energy and five others founded the Alliance for Critical Infrastructure back in February, built for the boring, unglamorous work of coordinating against cyberattacks and geopolitical disruption. What changed this summer is that AI became the reason the group needed to get much bigger, much faster, and organizers now expect the expanded version to be fully running before the year is out, working alongside the Trump administration on it.

---

Here's the part that's easy to flatten into a punchline and shouldn't be. At the same time Dimon is recruiting a safety coalition, his own bank is doing the opposite of slowing down. Employees in JPMorgan's asset and wealth management division now have access to something called LLM Suite, a generative AI tool built to do writing, idea generation, and document summarizing at something close to research-analyst level. A separate tool, Cash Flow Intelligence, lets corporate treasury departments forecast and analyze cash flows with it. More AI agents are reportedly coming later this year. This isn't a bank that's afraid of the technology. It's a bank that is deploying it aggressively and, in parallel, organizing forty other companies around the question of what happens when it goes wrong.

And Dimon's own public comments don't collapse into one clean position either, which is the actual interesting part. He's said his "anxiety is high" about how inflated AI-related asset valuations have gotten, that's a market-bubble worry, not a doomsday one. Separately, he's warned that AI has made cyber risk "much worse," because it hands the same capability upgrade to attackers that it hands to everyone else. But on the question most people actually lose sleep over, jobs, he's gone the other direction entirely, saying flatly that "we're kind of scaring the whole world much more rapidly than we should about it" and telling people to stop being breathless. So the same man is anxious about the money, anxious about the security, and actively unbothered about the thing dominating the public conversation. That's a much stranger position than "CEO says one thing, does another," and it's the one actually on the record.

None of this is happening in a vacuum where the worry is theoretical. Deloitte, one of the largest professional-services firms on earth, used generative AI to help produce a government report earlier this year and had to refund part of its fee after the report turned out to contain fabricated citations. That single incident was concrete enough that FINRA's 2026 regulatory oversight report added a dedicated section naming AI hallucination and bias as risks firms are now expected to test for and govern, not just discuss. A bank rolling out a tool built to do research-analyst-level writing is rolling it into the exact kind of work that just visibly failed somewhere else, in public, with a paper trail.

---

The part of this story that rarely makes it past the US business press is that JPMorgan's weight shows up just as plainly across the Atlantic, and not as a partnership announcement. Ahead of the UK's last Autumn Budget, after Dimon met personally with Chancellor Rachel Reeves, including a sit-down at Davos, representatives of the bank reportedly told her that JPMorgan would consider moving parts of its business out of Britain if the government raised the bank tax. That's not an AI story. It's a reminder of what kind of institution is doing the organizing here: one large enough that a single tax proposal in a G7 country becomes a conversation its CEO has directly with that country's finance minister.

Chancellor Rachel Reeves meets Jamie Dimon. Photo via HM Treasury, used under its Flickr terms (attribution required).

Meanwhile the UK's own banks are doing the opposite of what Dimon is doing at home: racing toward the same technology in public, as a selling point. NatWest became the first UK bank to partner directly with OpenAI this year, on top of a five-year AI and data modernization deal with AWS and Accenture, and hired Dr. Maja Pantic, formerly a generative AI research director at Meta, as its first chief AI research officer. Deutsche Bank and UBS are being ranked against each other on how well their AI strategy shows up in investment-banking league tables. Two different postures on the same technology, on two sides of the same ocean: Europe's banks are competing to look fastest, while the loudest name in American banking is privately building a shared brake pedal.

I don't think this is a story about hypocrisy, and I don't think it's a story about AI being secretly dangerous either. I think it's a story about proximity. Dimon runs the one institution in this whole picture with the clearest internal view of what the technology actually does when it's handling real money, real audits, real client data, not a demo of it. That view seems to be producing three different reactions in him at once, worry about the valuations, worry about the security, calm about the jobs, and none of those three reactions match the version of this conversation you'd get from reading headlines about either AI doom or AI hype. I don't know which of his three worries turns out to matter most. I'm not sure he does either. That's probably the most honest place to leave it.

FAQ

What is the Alliance for Critical Infrastructure, and is it new?

No. It was founded in February 2026 by nine companies, including JPMorgan Chase, Mastercard, AT&T, Lumen Technologies, AIG, Berkshire Hathaway Energy, Consolidated Edison, Southern Company, and Xcel Energy, originally to coordinate against cyber, physical, and geopolitical threats to critical infrastructure. In July, Jamie Dimon personally began recruiting more than 40 additional companies to expand it, repositioning AI as the group's central focus, with the expanded version expected to be fully operational before the end of the year.

Is Jamie Dimon against AI?

No, and JPMorgan's own actions show it. The bank has rolled out LLM Suite, a generative AI tool for its asset and wealth management staff, and Cash Flow Intelligence for corporate treasury forecasting, with more AI agents reportedly planned. Dimon has separately said AI will "change almost everything." His public concerns are specifically about inflated AI asset valuations and AI-amplified cybersecurity risk, not the technology's usefulness, and he has explicitly downplayed fears about AI-driven job losses as overblown.

What does this have to do with a real AI failure?

Deloitte, one of the world's largest professional-services firms, used generative AI to help produce a government report in 2025 that was later found to contain fabricated citations, and refunded part of its fee. The incident was significant enough that FINRA's 2026 regulatory oversight report added a section specifically naming AI hallucination and bias as risks financial firms must test for and govern.

Did JPMorgan really threaten to leave the UK over taxes?

According to reporting, yes. Ahead of the UK's last Autumn Budget, after Jamie Dimon met with Chancellor Rachel Reeves, including at the World Economic Forum in Davos, representatives of the bank indicated JPMorgan would consider relocating parts of its business out of Britain if the bank tax increased.

Are European banks investing in AI as heavily as JPMorgan?

Yes, and often more publicly. NatWest became the first UK bank to partner directly with OpenAI this year, alongside a five-year AI and data modernization deal with AWS and Accenture, and appointed a former Meta generative AI research director as its first chief AI research officer. Deutsche Bank and UBS are also being actively ranked against each other on AI-driven investment-banking performance in 2026 industry league tables.