Could US frontier AI labs dominate banking?
Tech.eu John Reynolds
AI labs are moving into finance, and banking is now part of the pitch. But most people at a fintech summit said they’ll partner with banks, not replace them.
Based on reporting by Tech.eu, John Reynolds — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Financial services is turning into one of AI’s most crowded arenas. Anthropic already says it gets more enterprise revenue from the sector than from any other industry except one, and OpenAI has hired investment banking experts to teach its models Wall Street workflows. That’s a clear sign the labs want in. The bigger question is whether they can move from useful tools to actual banking power.
At Accion Ventures’ Fintech for Inclusion Global Summit, the answer from the room was mostly no. Monica Engel of Quona Capital argued that the labs can get close to banking without becoming banks. Her point was simple: banking is slow, capital-intensive and tied to credit risk, which makes it a poor fit for a research lab. But she also said the money is still there — especially in agentic money flows, payments and commerce — and that the route in is partnership with regulated firms, not replacement.
Others drew a hard line between model quality and banking execution. Adrian Congiu of Mambu said frontier labs win when the product is the model itself: reasoning, text, code and images. Banking, he said, is different because the value sits in process knowledge, regulatory compliance, auditability and workflows stitched through old systems. He doubted any lab would build loan compliance for every geography, because that kind of breadth takes focus and the labs cannot go deep everywhere at once.
That theme kept coming back. Alexia Yannopoulos of Apis Partners said licensing, distribution and data still matter inside financial services. Ameya Upadhyay of Flourish Ventures recalled the old fear that Amazon would become the next big bank, then dismissed the idea of frontier labs doing the same, calling it too much of a headache. Rahil Rangwala of Accion Ventures put it bluntly: these companies want token usage at global scale, but regulation and local compliance remain strong moats for banking. The recurring conclusion was not that AI will stay out of finance. It’s that the labs are far more likely to sit on top of banking than to become it.
My take — AI-written commentary, not fact-checked reporting
The hype crowd loves to confuse access with ownership. In banking, that’s a lazy mistake: the real moat is compliance, distribution and the pain of being regulated, which is exactly why the labs will keep showing up as partners. The old dream of Big Tech or frontier AI simply becoming the bank always runs into the same wall: boring rules, and the rules usually win.
Read more about this at: Tech.eu