Trading floors became electronic exchanges over the past few decades. Bank branches gave way to apps, and cash gave way to a tap on a phone screen.
Each upgrade made finance faster without changing what sat underneath it: fragmented payment networks, multi-day settlement cycles, and reconciliation still done largely by hand.
That foundation is the part now being rebuilt. Payments, securities, compliance, and asset management are starting to run on connected digital infrastructure instead of separate systems stitched together by intermediaries.
Three forces are driving that shift at the same time: faster payment rails, AI embedded directly into bank operations, and a slower-moving effort to digitize ownership of traditional assets. Each is advancing on a different timeline.
How quickly real-time payments are actually growing
The most visible piece of this shift is happening in payment rails. The RTP network, operated by The Clearing House, now processes close to $500 billion in transaction value every quarter, according to American Banker.
The Federal Reserve‘s competing FedNow Service had grown to more than 1,700 participating financial institutions by April 2026, up from roughly 1,400 a year earlier, with a long-term target of expanding to 8,000.
The two systems do not yet talk to each other, which means banks running both rails maintain separate connections, fraud controls, and compliance processes for transactions that are functionally identical. That redundancy is one of the quieter costs of the current build-out.
Related: Fidelity cuts to the chase on 401(k) best practices
Growth is not limited to the U.S. PYMNTS data projects North American real-time payment volumes will reach 8 billion transactions in 2026 and nearly 13.9 billion by 2028, a 31.7% compound annual growth rate.
India’s UPI, the UK’s Faster Payment System, and Brazil’s Pix have already proven the model works at national scale, JPMorgan noted. The next phase, still mostly unsolved, is making these national systems talk to each other across borders.
Behind the payment rails, banks are leaning on cloud computing, APIs, and AI to handle fraud detection, compliance, customer service, and treasury management. The result is a system that looks less like separate institutions connected by middlemen and more like a single, data-driven network.
How AI is becoming embedded inside banking operations
Banks already use machine learning to flag fraudulent transactions, monitor financial crime, score credit risk, and manage liquidity. As industry data cited by Finzly confirm, these capabilities are taking on a growing share of operational work that used to require a person at every step, from settling payments to enforcing internal financial policy.
Treasury operations that once required a team checking multiple systems before releasing a payment are increasingly handled by software that checks compliance rules, available liquidity, and fraud signals simultaneously, then routes the transaction without a person touching it.
Porter Stowell, CEO of W3.io, argues the more important shift is that capital itself is becoming programmable, with AI systems handling execution while humans retain strategic decision-making.
“I think the biggest shift isn’t simply digitizing assets, it’s making capital programmable. Humans will continue making the strategic decisions, while AI handles everything else: moving capital, managing treasury operations, settling payments, and enforcing financial policies automatically,” Stowell said.
More Wall Street:
- HSBC doubles down on stock market message for 2026
- Citi quietly resets S&P 500 price target for the rest of 2026
- Jim Cramer has a stark message on the stock market for 2026
The scale problem is what he sees as most underestimated. Regulators built today’s oversight frameworks around human decision-makers acting at human speed. Software capable of acting autonomously and at high volume does not fit cleanly into those frameworks yet, and most regulators have not finished building the version that will.
“AI agents won’t execute twice as many transactions as humans, they’ll execute thousands of times more. That makes transparency, auditability, and governance far more important, because keeping track of autonomous capital movement becomes exponentially more complex,” Stowell added.
What’s behind the push to digitize traditional assets
A smaller but fast-growing piece of this shift involves representing traditional assets digitally on a shared ledger. The appeal is straightforward: Assets that have historically required large minimum investments and multi-day settlement can, in theory, be broken into smaller, more liquid pieces and settled almost instantly.
Ethra CEO Saeed Al-Marri argues that the bigger story is access, rather than the underlying technology.
“For the first time, ownership, settlement and compliance exist on the same digital infrastructure, eliminating layers of friction that have existed for decades,” he said.
His company launched a digital protocol in June, giving investors fractional access to maritime shipping vessels, an asset class where individual vessels typically run $30 million to $120 million and have historically required institutional-scale capital to access at all.
Private credit and government bonds have followed a similar pattern, with BlackRock’s digital money market fund alone passing $2 billion in assets under management.
Legal certainty remains the biggest open question, according to Al-Marri. Whether digital ownership records actually hold up when tested in court is a question most jurisdictions have not had to answer yet.
“Institutional adoption requires overcoming two barriers: legal settlement finality and regulatory fragmentation,” Al-Marri added. “Projects that tokenize institutional-grade, cash-flowing assets backed by absolute legal certainty will end up as the champions of real tokenization.”
Nd/Getty Images
The risk nobody is pricing into this transition yet
As AI systems start managing assets and executing trades independently, questions of identity, permissions, and accountability are becoming as important as transaction speed itself. Pharos CEO Wish Wu said future financial infrastructure needs to be designed for both human and AI participants from the start.
“As AI agents begin participating in financial markets, they’ll need infrastructure that is transparent, verifiable, and automated by design,” Wu said.
He raised a related concern about how the industry is framing risk in an automated system.
“As finance becomes increasingly automated, we also need to think beyond traditional cybersecurity. As AI agents start managing assets and executing transactions, questions around identity, permissions, and accountability become just as important as transaction speed,” Wu added.
The question of who is responsible when an autonomous system executes a bad trade, or when permissions are misconfigured across thousands of automated transactions, does not have a clean answer in most current regulatory structures.
Financial systems already run audit trails for human-initiated transactions, built around the assumption that a person made a decision and can explain it after the fact. Systems built around autonomous agents need a different kind of audit trail, one that can reconstruct which automated process decided to act and why.
What this buildout means for investors
The institutions building this infrastructure are positioned to shape the next phase of capital markets the way electronic exchanges and internet banking shaped the last one, more so than the institutions simply using it.
The competitive question shifting underneath the industry has moved beyond who controls capital. Now it is also about who builds and controls the rails through which capital moves.
For investors, that reframes where attention belongs. Faster payments and AI-driven banking operations are advancing at far greater scale right now than digital asset ownership models, even though the latter gets a disproportionate share of the headlines.
Whether any of this scales into durable infrastructure, rather than staying a collection of pilot programs waiting on regulatory frameworks to catch up, depends on the legal and compliance work happening underneath the surface.
The three pieces of this shift are not moving at the same speed.
- Payment rails are scaling fastest because the underlying problem has a clear regulatory path and obvious customer demand.
- AI inside banking operations is scaling almost as quickly because it solves a cost problem banks are already motivated to fix.
- Asset digitization is the slowest of the three because it runs into legal and jurisdictional questions that the other two simply do not face.
Investors tracking this space should separate those timelines rather than treating digital finance as a single trend moving at a single pace.
Related: Jim Cramer delivers strong buy call on fast-growing digital bank
