Real-Time Payments and AI in Commercial Payments

Real-time payments have yet to firmly establish themselves as a standard retail practice in the U.S., but significant transactions flowed through FedNow and RTP networks last year. Both systems recently raised their transaction limits to $10 million, greatly expanding potential business applications.

Increasing adoption of real-time payments will substantially reshape the B2B payment landscape. However, this is just one of several developments shaping a pivotal year for commercial payments.

According to Hugh Thomas, Lead Commercial and Enterprise Analyst at Javelin Strategy & Research, in his report on “2026 Commercial & Enterprise Trends,” artificial intelligence (AI)-driven automation and a shift towards more targeted pricing structures will play crucial roles.

An Inflection Year for AI

Optimizing payment flows—whether through automation or outsourcing—has always been a priority for finance leaders. Few technologies, though, can match the promise of AI.

Over the last few years, businesses across various industries have heavily invested in AI capabilities. This year will be decisive: organizations are now expecting clear returns on their investments.

Enhanced expectations come with the advent of agentic AI, which can significantly boost automation processes.

“You’re looking at a situation where so much work can now be automated,” Thomas said. “For example, upon initiating a purchase, you might start provisioning an agent to find goods or services that meet specific criteria—determine price points and check all necessary conditions before making the payment.”

“This year will likely see significant case studies emerge as more organizations implement AI,” Thomas added. “In accounts receivable, teams are already discussing how well-suited AI is to managing customer interactions on their AR portals.”

Historically, accounts receivable required constant human intervention—managing credit lines, reviewing invoices, reconciling payments, and handling exceptions. Generative and agentic AI can significantly reduce the time spent on these manual tasks.

Implementing AI securely and responsibly requires strong governance, oversight, and iterative deployment. Progress is likely to be incremental rather than sudden.

“I don’t know if we’ll see paradigm changes, but this year will mark a more common need for AI in payments,” Thomas said. “It’s still a learning phase, but many interesting case studies are expected.”

A New Real-Time Ballpark

In comparison to markets like India and Brazil, real-time payments have gained more cultural acceptance in the U.S., though domestic adoption is accelerating.

RTP, operated by The Clearing House, saw its transaction volume grow from 60 billion in Q2 2024 to approximately 481 billion in Q2 2025. FedNow, launched nearly three years ago by the Federal Reserve, has not displaced RTP but instead expanded alongside it, with FedNow facilitating around 246 billion payments in Q2 2025.

“You’re in a different realm now, where you see higher average value and clear use cases for instant fund transfers,” Thomas said. “A prominent example is housing down payments—moving from wire or cashier’s check to real-time payment, making both parties observe the funds moving between accounts.”

“This streamlines processes compared to handing a cashier’s check to a lawyer and waiting for confirmation of fund transfer,” he added.

Speed brings new risk management challenges, particularly fraud. Traditional payment systems had delays that allowed time for fraud screening and dispute resolution. With real-time settlement, these buffers disappear.

While instant payments introduce unique risk considerations, they also offer substantial benefits.

“These observable funds movements will drive quick adoption,” Thomas said. “They’ll drive the business case for managing new risk parameters and functionality that smaller banks will need to adopt at scale.”

Targeting Price-to-Value

As real-time payment networks gain momentum, card networks remain formidable competitors in B2B payments.

Credit card issuers have long aimed to replicate their consumer-market success in commercial payments. However, translating retail-based pricing models into the B2B environment has proven more complex than anticipated.

“There’s a lot of diversity among consumers, but not much difference in how they want to pay for things,” Thomas said. “People either seek rewards or access to credit, or they aim to minimize costs—and most know the best way to meet their own needs.”

“For example, when you go to a grocery store today, trying to pay with a check is rare; consumers can choose between card and cash. However, in business, organizations have multiple payment options like ACH, real-time payments, checks, direct debit, or cards,” he said. “There are many more choices here, often contingent on whether you prefer immediate payment or later settlements and the available discounts.”

B2B transactions operate under different economics, workflows, and value expectations, leading issuers to face well-established alternatives within enterprise finance teams.

Still, cards offer significant advantages in B2B contexts. Organizations can authorize one amount but settle for another with defined parameters, and chargeback rights provide strong recourse protections. From a control and risk-mitigation standpoint, cards remain one of the safest payment methods available.

To gain broader traction in commercial payments, however, issuers will likely need to move beyond retail pricing frameworks and adopt models aligned specifically to B2B value creation.

“The old Visa and Mastercard pricing documents used to be six or seven pages,” Thomas said. “Now, they’re around 30 pages, with many new sections describing different types of B2B transactions—pages for fleet payments, virtual card payments, and new card types.”

“The networks are getting smarter about pricing but lack visibility into the full transaction economics from both sides. They don’t know the costs and benefits seen by buyers and suppliers when using the network, nor do they see the rebates buyers might get or the costs to suppliers for accepting cards,” he said. “These new pricing schemes aim to balance transaction economics without controlling final costs, encouraging maximum use of their networks.”