QED’s Nigel Morris says AI will reorder finance from back office to banking

Nigel Morris, co-founder and managing partner of QED Investors and a co-founder of Capital One, says AI will be more consequential for financial services than the internet, mobile banking, cloud computing or blockchain. No funding round, valuation, revenue or headcount figures were disclosed; the significance is strategic, with Morris arguing that AI could compress operating costs across lending, compliance, service and back-office finance.

Morris has spent more than 40 years in financial services and helped launch Capital One in 1994, where he says the company used information-based strategy to reshape consumer lending. His central claim is that AI is different from prior technology cycles because it can affect every layer of financial services at once, rather than changing distribution or infrastructure alone.

Where Morris sees AI pressure first

Morris pointed to wealth management, investment banking, tax preparation, consumer banking, call centers, payments, expense management, identity and compliance as areas already being reworked by AI products. He cited Zocks in wealth management, Rogo and Model ML in investment banking, April in tax filing, and Decagon and Lorikeet in customer service.

He also named Ramp and Payhawk as companies pushing business finance toward more automated systems that combine cards, expenses, procurement and accounting. In risk and compliance, he cited Footprint and Sardine as examples of companies building for a market where identity verification, AML and KYC may need to account for transactions initiated by software agents.

Several of the companies Morris highlighted have ties to QED. Zocks, April, Albert, Lorikeet, Payhawk and Footprint are QED portfolio companies, according to Morris. Nubank was previously a QED investment, though the firm has exited its position after the company became public.

Cost compression is the core claim

Morris argues that AI’s first economic effect in finance is to push the marginal cost of work such as underwriting, compliance review and customer support closer to zero. That is a broad claim, and Morris did not provide cost benchmarks, adoption rates or productivity data to quantify it.

His view is that lower marginal costs could shift products from broad segmentation toward more individualized pricing and service. He described credit that adjusts with daily cash flow and insurance priced around individual risk rather than pooled averages as examples of products that become more plausible if AI systems can process data and make decisions at far lower cost.

For startups, Morris sees speed as the advantage. He argues fintech companies are better positioned to adopt AI quickly because they have shorter decision cycles and fewer legacy constraints. For incumbents, the advantage is data: large banks and insurers hold decades of transactions, balances, defaults and recoveries that startups cannot buy.

The unresolved issue, according to Morris, is whether incumbents can use that data aggressively enough. He cited earned-wage access, buy now, pay later, consumer-to-consumer remittances and digital brokerage as categories where established financial institutions gave up ground to fintech companies. He named Robinhood, Revolut, Stripe and Nubank as large fintech outcomes that followed.

Morris’s thesis is clear enough: AI is less a feature race than a cost and operating-model race. The weaker part is measurement. Without disclosed customer metrics, cost curves or adoption data for the examples cited, the argument rests on pattern recognition from prior fintech shifts and on QED’s view of where founders are building.

This story draws on original reporting from Crunchbase News.