Ramp’s recent $1.85 billion fundraise, coming just weeks after it secured $750 million in fresh capital, marks a second major wake-up call for banks lingering on the sidelines of the AI race, according to Malte Rau, co-founder and CEO of B2B payments and card-issuing platform Pliant.
Paired with Capital One’s $5.2 billion acquisition of Brex in April, the flurry of multi-billion-dollar activity from top-tier players is forcing traditional financial institutions to confront a priority they had long kept on the back burner.
While the largest twenty banks were already issuing their share of AI-related RFPs, Ramp’s latest announcement threatens to kick bank interest into overdrive.
“Banks are waking up to this,” Rau said. “They say they have to take this seriously because a top-10 bank bought a solution, and Ramp is really bringing the pain. They have a lot of questions.”
Photo by Julio Lopez on Unsplash
But beneath that urgency, institutional hesitation remains.
“Effectively, they want to buy a 2022 version of Ramp,” Rau said. “They would love to play in the AI game, but because of pure data security or info security, they wouldn’t be allowed to onboard AI for something like that.”
Many institutions are merely dabbling in AI while biding their time for regulatory clarity and security concerns to resolve. Potential bank clients want the functional benefits of AI without having to mention AI. With onboarding times of close to a year, they hope that more basic AI will suffice.
Rau estimates that older iterations of AI will satisfy 99% of the market for now. With adoption still in its earliest phase—perhaps 2,000 out of a 70,000-strong market actively deploying solutions—the vast majority of market share remains up for grabs.
Market response has also split along enterprise lines. Providers servicing small and mid-sized businesses have remained mostly uninvolved. At the mid-market and enterprise levels, though, complex internal approval flows have spurred larger players to seize opportunities.
Rau said there is a large discrepancy between what features providers can offer and which ones clients will actually use. Consumers want to order a card with a button press, or get a mobile phone receipt rather than advanced backend capabilities.
“They’re not talking about zero-day close,” Rau said.
That shift is moving along an unconventional path, according to Rau. While technological innovation typically gains a foothold in B2C before migrating to B2B, enterprise finance is proving to be a rare counterexample. AI is already showing immediate utility in onboarding, compliance, and automated risk underwriting, where autonomous agents underwrite more accurately than human reviewers, he said.
The convergence of agentic finance and commercial cards is granting legacy card rails a second life. In the complex landscape of B2B transactions, cards can be issued directly to autonomous agents programmed with deterministic spending controls.
“More and more want to directly connect the AI provider with the card or payment infrastructure,” Rau said.
The United States represents a prime market for agentic payments, where autonomous software agents can select optimal payment rails, verify counterparty data, and reconcile transaction records. AI-assisted card operations also provide a dual economic benefit: unlocking high-margin interchange revenues while cutting down operational overhead.
“Banks always knew this was a big opportunity, but now they’re realizing this is also cheaper than having someone pick up a phone,” Rau said.
Still, the pace of enterprise sophistication will likely remain muted over the next several years. While data-driven and e-commerce companies run ahead with deeper technical integrations, the broader enterprise market will continue to prioritize personal relationships alongside machine efficiency.
“People want to buy from other people,” Rau said. “Even if they explain AI, especially enterprise, they want to have a person on the other end.”
