This capability is enabled by a new collaboration between WEX, which is a provider of intelligent payment solutions, and Extend, an embedded virtual card payments company that is an established SAP Concur partner, the companies said in a Thursday (April 30) press release.

WEX corporate card customers can initiate these vendor payments by connecting their WEX commercial account in Concur Invoice, according to the release.

Then, when an invoice is received, Concur Invoice automatically generates a virtual card linked to the customer’s registered WEX commercial account and authorizes, remits and reconciles the payment.

This solution enables WEX customers to settle invoices with single-use virtual cards generated from their WEX commercial account, control payment amounts and timing, pay suppliers faster, automate reconciliation, gain complete visibility into payment delivery, and earn available card rebates on vendor transactions.

“By partnering with Extend to embed virtual card payments inside Concur Invoice, we are delivering infrastructure that offers granular control over every transaction, all without disrupting how they already operate,” Carlos Carriedo, chief operating officer, Americas Payments & Mobility at WEX, said in the release.

The collaboration demonstrates Extend’s ability to help corporate payments providers and financial institutions deliver embedded payment capabilities inside the platforms their customers rely on, without replacing what’s already working, per the release.

“As customer expectations rapidly evolve, so does the race to deliver more connected payment workflows — Extend is excited to be the partner that makes it possible,” Extend CEO and Co-founder Andrew Jamison said in the release.

The PYMNTS Intelligence report “FinTechs Tap Embedded Payments to Deepen Customer Relationships” found that among the firms that have adopted embedded payments, 87% cite better customer experiences as a key benefit, while 60% cite greater consumer trust and 53% cite improved operational efficiency.

Another PYMNTS Intelligence report, “Embedding Security: Designing Fraud Risk Out of Business Transactions,” found that 74% of embedded finance users say the technology has significantly reduced their fraud risk, suggesting that well-designed embedded payment systems can outperform traditional standalone fraud monitoring tools.

What’s changing isn’t simply the technology used to move money—it’s who is shaping the experience and setting the rules. Beneath B2B payments’ long‑standing inertia, a structural shift is underway that is redefining control, acceptance and value across the transaction lifecycle.

To explore that shift, PYMNTS sat down recently with Mastercard Senior Vice President of Commercial Acceptance Nick White and Wells Fargo Head of Commercial and Corporate Banking Merchant Services Paul Uher.

“Two-thirds of B2B suppliers report that they’re not meeting their buyers’ expectations for their payment experience,” said White, noting that the total addressable opportunity in B2B payments is around $80 trillion.

The significance of that opportunity is not simply its size. It is that the benefits across the market appear to be aligning. Buyers want greater control, security and working-capital flexibility. Suppliers want faster payment, lower administrative burden and fewer late receivables.

Turning to cards as a B2B payment mechanism can increasingly offer them those advantages.

“Nearly half of [suppliers] are saying that they expect to be asked to take card for a B2B payment,” White said.

The evolving role of cards reflects a broader realization across the B2B payments industry: scaling card acceptance in commercial payments is less about technology alone and more about solving entrenched operational friction.

It’s against that backdrop that acquirers are emerging as crucial strategic orchestrators of B2B payment ecosystems.

From Closing the Acceptance Gap to Enabling Seamless Coordination

Virtual cards are often positioned as a silver bullet for B2B payments, but their real value lies in how they reconcile competing priorities. For buyers, they offer granular control over spending, improved cash flow management and potential rebates. For suppliers, they provide faster payments, reduced credit risk and streamlined processes.

The result is a fundamental shift in B2B acceptance, where the focus moves beyond a persuasion process and toward a broader question of enablement: who can make it easier for enterprise buyers and suppliers to transact at scale with less friction, better data, faster reconciliation and more flexible payment choice.

“The suppliers are a lot more in the driver’s seat now for these discussions, as it relates to the operational challenges they’re facing,” Uher said.

One key area of opportunity, he added, is optimizing how virtual cards are integrated into existing workflows—ensuring payments and remittance data flow seamlessly, minimizing manual touchpoints and accelerating reconciliation.

“The last thing they need is human beings involved,” White said, explaining that virtual‑card‑enabled payments remove manual touchpoints from invoice processing, enabling suppliers to operate at scale with faster settlement, cleaner data and more reliable reconciliation.

“With straight-through processing and reconciliation, we can take the payment instruction, settle the payment and move the reconciliation information with no human intervention,” he added.

That is why ecosystem connectivity is emerging as a defining theme across B2B. The opportunity is no longer just to sign up more suppliers. It is to connect the right counterparties, route the right transactions through the right rails, move both money and information together and do so in a way that does not create new manual burdens.

The Acquirer’s Strategic Role in B2B’s Upgrade

If supplier operations are now the battleground, the acquirer’s role necessarily changes. In consumer commerce, card acceptance is table stakes. In B2B, it is still consultative. Acquirers have to do more than enable transaction processing and coordinate a complex commercial system that includes issuers, networks, suppliers, buyers, treasury functions, ERP environments and, increasingly, AI-driven decision engines.

“You need the acquirers. The acquirers have to be in place and have the capabilities in place,” Uher said.

These capabilities are not glamorous, but they are foundational: sales teams that understand the nuances of B2B acceptance, ERP file integration, automated posting, invoice presentment, straight-through processing and the ability to route payments intelligently across different rails. Without those basics, the promise of card-based B2B payments may remain trapped in pilot mode.

Both Uher and White emphasized that suppliers are not monolithic organizations with a single decision-maker. Treasury may focus on working capital, while finance operations may care about reconciliation and labor cost. Providers now need a multithreaded sales motion that can speak to each constituency inside the supplier.

“It is no longer enough just to talk to the person who’s signing off the bill for a payment and acceptance,” White said.

AI Moves From Hype to Real-World Value

No contemporary payments discussion avoids artificial intelligence (AI), and both White and Uher positioned the technology as a practical tool for reducing friction across the supplier lifecycle—applying intelligence to areas where it can drive real operational impact.

For White, AI’s first major use case was intelligence: identifying which suppliers are likely to accept cards, shaping more relevant sales narratives, and helping automate outreach, onboarding and optimization at scale.

“The opportunities [are] endless,” he said, though always in service of a simple goal: “to solve problems, to automate, to scale.”

Uher noted that AI’s biggest immediate leverage may lie upstream, in decision-making rather than payment execution. He pointed to its potential to help suppliers determine “where to route transactions,” deciding when card makes sense versus ACH, wire, or another payment type.

Ultimately, if the two payments executives came to a shared conclusion, it was that the market has moved beyond the question of whether B2B card acceptance matters. The question now is who can operationalize it.

White called the current moment “an inflection point,” citing stronger buyer demand, greater supplier openness and a growing recognition among acquirers that B2B acceptance can deepen merchant relationships while creating new growth.

Uher’s advice to suppliers was pointed in the same direction: engage your current provider, understand what capabilities you are not yet using and plan for the technology and workflow changes needed to make acceptance easier.

For all PYMNTS B2B coverage, subscribe to the daily B2B Newsletter.

The gap between collecting data and extracting advantage from it has become one of the defining competitive fault lines in modern business.

“Most companies are sitting on an enormous amount of data today, and they’re doing very little with it,” Rinku Sharma, chief technology officer at Boost Payment Solutions, told PYMNTS during a discussion for the April edition of the “What’s Next in Payments” series, “The Data Game.”

“The moat is not having more data. The moat is about having the right data, knowing what questions you want to ask of that data, and building the infrastructure to act on those answers at speed,” Sharma said.

Nowhere is this shift more visible than in B2B payments, a sector historically defined by fragmentation, manual processes and limited visibility. But by connecting data across products, networks and workflows, firms that can harness the ecosystem’s latent intelligence are finding themselves able to provide a coherent view of financial activity that informs strategy as much as execution.

“The companies that are winning today are not the ones who have waited for everything to be perfect, but the ones who have started building with what they have and iterated relentlessly,” Sharma said.

Boost’s own ambition, he added, is to unlock the ability of data to answer higher-order questions for clients, such as where working capital risks are emerging, how supplier networks behave and where performance can be optimized.

From Data Exhaust to Intelligence Layer

In consumer payments, innovation is often measured in visible touchpoints: checkout speed, interface design or approval rates. B2B payments operate differently. The “experience” is embedded deep within enterprise systems such as ERP platforms, accounts payable workflows, treasury infrastructure and more, making it both less visible and historically slower to evolve.

“What we’re doing now is building the analytical layer that turns these signals into intelligence,” Sharma said. The goal is not retrospective reporting, but operational impact, “faster decisions, smarter interchange optimization, and better outcomes for our buyers and suppliers.”

Because as industries digitize, competitive advantage is being shifted away from ownership of infrastructure toward ownership of insight.

Operating at the intersection of buyers, suppliers and card networks, Boost processes commercial card transactions across thousands of relationships in more than 180 countries. That vantage point yields a dataset no single participant could assemble independently, but raw visibility is only the starting point.

“It’s not magic,” Sharma said. “It’s a result of data that is extracted from payments and about the supplier’s systems being understood and acted upon automatically.”

The impact can be tangible. Suppliers joining a modern payments network, Sharma said, increasingly receive funds automatically, accompanied by clean, structured remittance data in their preferred formats.

The elimination of manual reconciliation is not the result of a single feature, but of coordinated intelligence across systems capable of understanding supplier preferences, buyer workflows and transaction context simultaneously.

AI Collapses Distance Between Signal and Action

Data intelligence’s impact across B2B is also getting a concurrent boost from artificial intelligence, which is compressing the time, cost and complexity required to extract value from data.

Sharma framed AI’s impact along three dimensions: speed, cost and quality, and argued that its real power lies in how those forces compound.

“Transaction decisions that were happening historically in manual review queues or overnight batch jobs are now happening in milliseconds,” he said, noting that for large-scale B2B payments, that means parsing unstructured data, optimizing interchange and delivering enriched remittance information instantaneously.

The reason this is now possible is that cost dynamics have shifted just as dramatically. What previously required “an entire team of data scientists running and building ML [machine learning] models at scale” — along with significant compute investment — can now be deployed “in days and weeks without that huge upfront investment,” Sharma said.

But it is in quality where AI introduces the most profound change. Traditional systems have relied on predefined rules, inherently limited by what engineers anticipate. Machine learning, by contrast, surfaces patterns that were never explicitly encoded, enabling detection of anomalies and opportunities that would otherwise remain invisible.

“It looks for patterns that it has never been coded for,” Sharma said.

The result is not simply better automation, but a fundamental expansion of what organizations can perceive and therefore act upon. By analyzing cash flow patterns, payment velocity and network relationships, companies can make faster, more precise decisions.

“What real-time transaction data is doing is enabling us to have a forward-looking assessment,” Sharma said. “The question used to be what happened. Now the question is, what should we do about it right now?”

In that sense, the future of data-driven advantage may be less about ownership than orchestration. Companies that can unify disparate signals, interpret them in real time and act with clarity may define the next generation of competitive moats.

For Boost, that means evolving beyond payments infrastructure into what Sharma described as “the intelligence layer in B2B commercial payments.”

“When a supplier asks why a transaction was routed in a certain way, they deserve an answer,” he said. “The companies pulling ahead are using AI and payments as a trust-building and growth engine.”

For all PYMNTS B2B coverage, subscribe to the daily B2B Newsletter.

The new agents can triage employee requests, source vendors, review contract terms and handle compliance checks, the company said in a Wednesday (April 29) press release.

Because the platform runs on anonymized pricing benchmarks and vendor data from millions of Ramp transactions, it can provide smaller companies with the sort of benchmark data that might be used by Fortune 500 firms, according to the release.

The company’s new AI agents reinforce this effort with their abilities to spot compliance risks and identify savings opportunities, per the release.

The need for these sorts of capabilities has risen with the growing adoption of AI by businesses, because AI contracts are increasingly expensive and feature new kinds of pricing models, the release said.

“The tools companies use to buy haven’t kept pace with the speed or sophistication of what they’re buying,” Ramp Chief Product Officer Geoff Charles said in the release. “We built a purchasing platform where AI agents do the work. Finance teams can hire Ramp as an extension of their team to run purchasing end-to-end.”

Ramp Procurement customers, on average, reduce their vendor costs by 16% annually and cut the amount of time spent on manual purchasing work by 46 hours per month, per the release.

The PYMNTS Intelligence report “The Investment Impact of GenAI Operating Standards on Enterprise Adoption” found that among U.S. firms generating at least $1 billion in annual revenue, 25% are actively using generative AI in their procure-to-pay cycle and another 48% are considering doing so.

Ramp released its first set of AI agents, Agents for Controllers, in July 2025 and its second, Agents for AP, in October 2025.

In March, Ramp and Visa announced that they were introducing AI agents that automate corporate bill pay, reduce manual work, curb spend and unlock savings. The companies said these agents would provide Ramp customers with greater payment flexibility and more control over corporate spend.

Ramp said at the time that its enterprise customer base grew 133% year over year in 2025 as companies seek a replacement for old infrastructure burdened by manual controls and disconnected systems.

For all PYMNTS B2B coverage, subscribe to the daily B2B Newsletter.

The portal is built to process up to $127 billion in tariff refunds the U.S. owes businesses, offering structured fields, standardized uploads and a streamlined process designed to replace years of opaque, paper-heavy compliance.

CBP said importers and authorized customs brokers can now use their Automated Commercial Environment Secure Data Portal (ACE Portal) accounts to file Consolidated Administration and Processing of Entries (CAPE) Declarations for International Emergency Economic Powers Act (IEEPA) refunds.

But for the more than 330,000 importers that paid tariffs, the challenge was never the login. It was being ready for what the portal would ask of them.

The firms that moved fastest were not necessarily those with the largest refunds at stake, but those that had already done the invisible work of structuring their data, aligning internal processes and maintaining operational records in a structure and richness that could meet the portal’s requirements.

The lesson here is not about tariffs. It is about how modern B2B systems operate. Whether the touchpoint is a government portal, a marketplace, or a payment system, it functions as the final checkpoint in a much larger operational pipeline.

What determines success across that pipeline is everything that happens before anyone ever clicks “submit.”

See also: It’s Level 3 or Bust as Visa’s Interchange Shift Rewires B2B Data 

Data Readiness as Compliance Function

In practice, the CBP’s tariff refund portal functions as a validation engine, enforcing strict requirements on data accuracy, formatting and completeness. It assumes that the submitting organization has already reconciled its internal records to a high degree of precision.

In that sense, it resembles a growing class of B2B interfaces that serve as gateways rather than workspaces. They do not help organizations create order; they require that order already exist.

One of the clearest takeaways from the portal rollout is the cost of fragmented systems. Many companies operate with a patchwork of tools — ERP systems for finance, specialized platforms for logistics, custom databases for product information. Each system evolves independently, optimized for local efficiency rather than global coherence.

That fragmentation can become a liability when a unified view is required. In the context of tariff refunds, this meant reconciling shipment data with tariff classifications, payment records and historical filings. Discrepancies that had been tolerable in day-to-day operations could suddenly became blockers.

These problems and hurdles are not merely technical; they are also organizational. Data fragmentation often mirrors organizational silos, where different teams operate with their own systems, priorities and definitions.

PYMNTS Intelligence found in December, for example, that 66% of accounts payable teams saw an increase in manual workload over the prior year.

In this environment, the competitive advantage shifts decisively away from the interface and toward preparation. Companies that succeed are not those that navigate the portal most efficiently, but those that arrive at it with the cleanest, most coherent datasets and the clearest internal alignment.

This reframes what “enablement” means. It is no longer sufficient to provide users with access and instructions. True enablement lies in equipping them to meet the system’s implicit demands: data integrity, cross-functional coordination and anticipatory compliance.

New data in the “2025–2026 Growth Corporates Working Capital Index: North America Edition,” a collaboration between PYMNTS Intelligence and Visa, reveals a widening performance gap between firms that have modernized their receivables and working capital infrastructure and those that continue to rely on manual, legacy processes.

See also: Cross-Border Payments Hit a New Bottleneck at the Data Border

Workflow Tooling as Infrastructure

Walmart, Target and Nike are reportedly expecting some of the biggest refunds, owed $10.2 billion, $2.2 billion and $1 billion, respectively. FedEx and DHL have started filing claims, while UPS has said it will work to request and retrieve refunds from the government on customers’ behalf for shipments where UPS was the company of record.

By the time 2025 was winding down, 47% of goods product leaders surveyed by PYMNTS said tariffs were mostly or completely negative for their company finances, while 88% were still anticipating supply chain disruptions. At the same time, around two-thirds of goods firms and 80% of services firms said tariffs could eventually bolster supply chain resilience.

And if the CBP refund portal is the last mile, it also shows how competitive advantage is shifting upstream. The organizations that succeed are those that invest in what might be called pre-submission intelligence — the ability to understand, structure and validate their data before it encounters an external system.

This can require greater visibility across workflows, the ability to trace how information moves through an organization, and mechanisms for identifying inconsistencies early. It may also require a cultural shift, where data is treated not as a byproduct of operations but as a strategic asset.

Data readiness and ensuring that information is accurate, structured, and accessible at all times has effectively become a new compliance function. Every transaction, every update, every classification decision contributes to a dataset that may eventually be tested against an external system. By the time that test occurs, it is too late to fix foundational issues without significant cost.

After all, in a world where clicking “pay” is the easy part, the real work and the real differentiation increasingly now lies in everything else that comes before.

Findings in a new report, “The Emerging Middle Market: When Operational Complexity Grows Faster Than Financial Infrastructure,” a collaboration between PYMNTS Intelligence and i2c, show that 46% of high-growth mid-market firms surveyed report frequently missing opportunities because available credit is too slow, too rigid or simply misaligned with their needs.

Businesses generating between $10 million and $50 million in annual revenue increasingly find themselves stranded between two worlds. They have outgrown entry-level banking products, basic accounting tools and single-provider payment stacks. But they remain too small, or too early in their profitability trajectory, to justify enterprise-grade treasury systems or bespoke financial relationships. The financial infrastructure around them hasn’t evolved at the same pace.

The result is a widening gap between what these companies need and what financial institutions are equipped to deliver.

This reality is forcing even the fastest-growing companies to manage cash on spreadsheets, defer investment and navigate fragmented payment and credit systems.

The Fastest-Growing Companies Have the Least Financial Visibility

One of the clearest manifestations of this gap is in cash visibility. The report data shows that firms growing at the fastest rates — those expanding at 20% annually or more — are also the most likely to experience frequent cash-flow disruptions.

For many, these disruptions occur weekly or even daily, at rates several times higher than slower-growing peers. This is not simply a function of size; smaller firms with steady growth profiles often report fewer issues. Instead, liquidity stress correlates directly with the pace of expansion.

The reason? The data shows that, even as these businesses expand rapidly, the financial infrastructure supporting them hasn’t kept pace, leaving them without integrated, real-time visibility across payments, accounting and liquidity.

In effect, the very firms driving economic momentum are doing so with the least financial control, not because they lack sophistication, but because the systems around them haven’t evolved to match their speed.

And if visibility is one side of the problem, credit is the other. On paper, most emerging middle-market firms report adequate or even more-than-sufficient access to credit. But this headline metric obscures a deeper disconnect.

The Problem in Credit Deployment

Among the fastest-growing firms, nearly half say they frequently miss opportunities due to insufficient financing — not because credit isn’t available, but because the products offered are too slow, too rigid or too disconnected from real-time business needs.

For companies scaling rapidly, traditional underwriting metrics often fail to capture future potential or current demand. As a result, credit facilities tend to be too slow to approve, too rigid to adjust and too limited to support real-time decision-making.

Read the report: The Emerging Middle Market: When Operational Complexity Grows Faster Than Financial Infrastructure

But if the scale of the gap is significant, so too is the opportunity for financial institutions willing to evolve their infrastructure and product offerings. As more companies enter the emerging middle market, demand is shifting toward integrated, real-time financial infrastructure that aligns more closely with how these businesses operate.

For example, systems that link payment data with accounting records and credit availability can allow a company to instantly assess its capacity to take on a new project or expand inventory. At the same time, automated reconciliation can reduce the labor hours spent on manual processes, freeing finance teams to focus on analysis and planning. Simultaneously, dynamic credit solutions informed by real-time data can provide more flexible and responsive funding options.

After all, if the problem is fragmentation and latency, the solution is integration and immediacy. The next generation of financial infrastructure must provide a unified view of payments, accounting and liquidity, one that is updated in real time and accessible across the organization.

For financial institutions, the implication is clear:

• Client growth is accelerating — but financial infrastructure and product offerings are not keeping pace.
• Closing that gap is not just a technology challenge — it’s a strategic opportunity to better serve one of the fastest-growing and most underserved segments in the market.

For the companies themselves, the stakes are high. In a competitive landscape where speed and agility are paramount, the ability to see, decide and act in real time may be the difference between leading the market and watching opportunities pass by.

As of Saturday (April 18), Visa’s Commercial Enhanced Data Program (CEDP) officially phased out its Level 2 interchange program, replacing the legacy incentive structures with a new interchange system that rewards one thing above all: high-quality, verifiable Level 3 data.

What was once an optional optimization for B2B card payments is now the primary pathway to cost efficiency. For CFOs and operators, the implications may be as immediate as the Level 2 sunset, and their consequences are not merely technical ones.

Transactions that previously qualified for discounted rates under Level 2 will now fall into higher-cost categories unless they meet stricter Level 3 standards. There is no longer a low-effort path to optimization. Transactions either meet the bar for verified, high-quality data, or they don’t. That distinction carries real financial weight.

In the end, the shift is less about payments than about preparedness. The companies that succeed will be those that recognize the moment not as a disruption to be managed, but as a catalyst for transformation.

Read more: Why Messy Merchant Data Could Make B2B Payments More Expensive 

A Structural Reset in Interchange Economics 

In a world where data drives value, the ability to capture, manage and leverage detailed transaction information becomes a defining capability.

CEDP does not simply require more data; it demands better data. Visa now validates transactions for completeness and accuracy, assigning merchants a “verified” or “non-verified” status that directly determines interchange rates.

Level 3 data has long been associated with large-ticket or specialized transactions, requiring detailed line-item information such as product descriptions, quantities, unit costs and freight charges. What has changed is its centrality. With Level 2 removed, Level 3—or more precisely, CEDP-validated Level 3—is now the only scalable way to achieve interchange savings in B2B card payments.

Legacy payment architectures are rarely designed for this level of precision. Data often originates in disconnected systems spanning enterprise resource planning (ERP) platforms, invoicing tools, procurement workflows and more, and must be stitched together before a transaction is processed. Each handoff introduces risk: missing fields, inconsistent formatting or incomplete records that can invalidate an otherwise eligible transaction.

Zach Lynn, head of customer data and insights at Boost Payment Solutions, wrote in the PYMNTS eBook “Headlines That Will Shape the Close of 2025” that data exchange has become non-negotiable in the payments industry.

“In today’s environment, seamless, secure data flows between buyers, suppliers and financial institutions are essential,” Lynn wrote. “Whether it is enabling real-time reconciliation or supporting advanced analytics, the ability to move and leverage data is now table stakes for any organization serious about optimizing working capital.”

Because Visa now evaluates data quality transaction by transaction, small errors can cascade into higher costs. Inaccurate data doesn’t just fail quietly; it actively erodes margin.

The result is a shift in priorities. Finance teams that once focused on negotiating processor rates must now invest in data governance, system integration and automation. Payment optimization becomes as much a technology problem as a financial one.

See also: The Riskiest Words in B2B: This Is How We’ve Always Done It 

From Cost Center to Strategic Lever

One of the less discussed implications of the Level 2 sunset is its effect on financial planning. Interchange has always been a variable cost, but it was also relatively predictable. Finance teams could model expenses based on historical qualification rates, assuming a stable mix of Level 1, Level 2 and Level 3 transactions. That assumption no longer holds.

Under CEDP, interchange becomes more dynamic. Rates fluctuate based on data quality, verification status and ongoing compliance. Transactions can be downgraded retroactively if data fails validation.

This introduces a new layer of volatility into forecasting. CFOs must now model not just transaction volume, but data performance—estimating how much of their payment flow will qualify for verified status versus standard rates.

For businesses that rise to the challenge, the rewards are tangible: lower costs, greater efficiency and a more scalable operating model. For those that do not, the consequences will be equally clear, reflected in higher fees and missed opportunities.

Looking ahead, if firms persist without the systems and processes to support high-quality data, they may face persistently higher costs and reduced flexibility particularly as concurrent innovations like artificial intelligence scale across the B2B space.

“If you don’t have well understood, well managed, well governed data, it’s going to be really hard to use AI,” Kathleen Pierce-Gilmore, senior vice president and global head of issuing solutions at Visa, told PYMNTS this week.

“I would put my whole life savings on the modernization of infrastructure,” Pierce-Gilmore added.

The future may not be as far off as some CFOs believe. On Friday (April 17), for example, billing and revenue automation platform Zenskar raised $15 million in new funding to expand its agentic artificial intelligence (AI) capabilities.

As AI agents become more capable, the organizations that can feed them high-quality, structured data could be better positioned to gain a decisive advantage over those that cling to document-centric and human-optimized processes.

The problem is that most B2B contracts and AP/AR workflows were never designed for an agentic reality. These fundamental operational artifacts are frequently dense, ambiguous and optimized for human judgment rather than machine execution.

In contrast, an AI-native organization would treat traditional contracts and invoices as structured data from the outset. Every clause, payment term and obligation becomes a discrete, machine-readable element. Instead of parsing text after the fact, systems can directly ingest, interpret and act on the data in real time.

A contract written for humans requires interpretation. A contract structured for machines enables execution. In this emerging landscape, B2B contracts and invoices are no longer just legal documents but operational instructions that data management best practices make interpretable by machines.

See also: What Agentic Commerce Can Learn From B2B Payments

Does AI Spell the End of Human-Only B2B Contracts?

With contracts and invoices becoming machine-readable, the possibilities of fully autonomous financial workflows can come into focus for CFOs. AI agents can negotiate terms within predefined parameters, execute transactions based on real-time conditions and enforce agreements without human intervention.

For CFOs, this represents an opportunity to reduce cycle times, improve cash flow visibility and minimize disputes. But it also requires a willingness to rethink long-standing processes and invest in data infrastructure.

If contracts are to be executed by AI agents, they must be designed with machine readability in mind. Key elements such as pricing models, payment schedules, service-level agreements and termination conditions should be explicitly defined in standardized formats. Ambiguity, while sometimes useful in human negotiations, becomes a liability when machines are tasked with execution.

Invoices are undergoing a similar transformation. In an AI-driven environment, invoices become real-time signals within a continuous financial workflow. They are generated, validated and reconciled automatically, with discrepancies flagged and resolved by agents before they escalate.

To enable this, invoices must be standardized and structured from the point of creation. This includes consistent data fields, clear linkage to underlying contracts, and integration with procurement and payment systems.

Findings from PYMNTS Intelligence’s November edition of the Payments Optimization Tracker® Series revealed that as agentic AI systems mature, descriptions optimized for human persuasion, like rich imagery, narrative copy and lifestyle framing, must be complemented by precise, unambiguous metadata, like specifications, dimensions, compatibility, warranties, return policies and availability in consistent formats.

See also: Agentic AI Puts a Face on Corporate Treasury’s Next Leap

How CFOs Can Structure Their Organizational Data for the AI Era

The transition to agent-ready contracts and invoices is not a single project; it is a foundational shift in how organizations think about data. It requires investment in systems, alignment across functions and a willingness to challenge entrenched practices.

Research from PYMNTS Intelligence has shown that 83% of companies have yet to fully automate their accounts receivable operations, with data fragmentation a key limitation.

For CFOs, the starting point is often an audit of existing processes and data structures. Where are the inconsistencies? Which documents are most critical to operations? How easily can key terms be extracted and acted upon?

Billtrust Chief Product Officer Lee An Schommer said in a recent PYMNTS interview that companies manage an average of three enterprise resource planning (ERP) systems, leading to data silos that make it tough to get a unified view of customer behavior, payment history and dispute patterns.

At the same time, the introduction of AI agents into contractual and financial workflows is likely to raise important questions about accountability. If an agent negotiates a term or triggers a payment, who is responsible for the outcome?

Organizations must also look to remain vigilant about the risks associated with automation. Bias in training data, errors in logic and unforeseen edge cases can all have material consequences. Robust oversight mechanisms are essential to mitigate these risks.

But while it can be tempting to view the agentic transformation as a matter of adopting new tools or platforms, in reality, it is a competitive reset. In a world where machines are becoming active participants in commerce, the structure of contracts and invoices may determine which firms are able to move faster, operate more efficiently and respond more effectively to changing conditions.

That calculus has not disappeared. It has migrated. Today’s buyers may not be defaulting to a single blue-chip vendor, but they still gravitate toward incumbents, familiar workflows and systems already embedded in organizational muscle memory.

Nowhere is that instinct more visible than in B2B payments. Physical checks remain stubbornly present — and for understandable reasons. They offer visibility, control and auditability. They map to established approval hierarchies, satisfy compliance workflows and distribute accountability across accounts payable, finance and management, diffusing the risk any single decision-maker must absorb.

Digital B2B payment alternatives, by contrast, can require organizations to trust new mechanisms and, in some cases, relinquish familiar safeguards.

Change is scary. But the “safe” dynamics of traditional payment methods are increasingly being tested by a new operational reality in which financial experiences are central to growth.

As automated payments, embedded finance solutions and dynamic credit systems scale across the marketplace, the old maxim may be due for a replacement: No one ever lost business by making it easier for their customers to pay.

See also: How CFOs Are Turning B2B Payments Into a Strategic Weapon  

Why Business Growth Begins Where Money Moves

For decades, payments in B2B were treated as an operational afterthought. Invoicing, net terms and reconciliation processes were viewed as back-office concerns, largely disconnected from customer acquisition or retention.

The result was a paradox, where companies invested heavily in modernizing customer-facing experiences while maintaining antiquated payment infrastructures that introduced friction for both customers and internal teams alike.

PYMNTS Intelligence found in December, for example, that 66% of accounts payable teams saw an increase in manual workload over the prior year.

But in today’s environment, the payment layer is often the first meaningful interaction a customer has with a vendor’s system of trust. Before a product is fully adopted, before integration is complete, customers must evaluate how they will pay, when they will pay and what risks they assume in doing so. These considerations are no longer trivial, they are becoming gating functions for growth.

“The office of the CFO is broadening its mandate,” Boost Payment Solutions founder and CEO Dean M. Leavitt told PYMNTS earlier this year, adding that while decisions about how companies pay and are paid “were traditionally a secondary issue for most CFOs,” that legacy hierarchy is now changing as finance leaders come to recognize the working capital implications of strategic B2B payment design.

A company offering frictionless onboarding but rigid payment terms may struggle to convert, while a platform with superior capabilities but opaque billing structures could lose to a competitor with clearer financial workflows. Even subtle factors such as whether credit limits can scale dynamically or whether fraud controls create false positives can now determine whether a deal progresses or stalls.

Read also: CFOs Ditch AI Features to Fix Broken Payment Flows 

Switching Costs and the Psychology of Change

If the old IBM adage was about avoiding blame, the emerging paradigm may be about redefining what constitutes a defensible decision. In a world where growth increasingly starts at the payment layer, clinging to outdated systems can be as risky as adopting unproven ones.

New data in the “2025–2026 Growth Corporates Working Capital Index: North America Edition,” a collaboration between PYMNTS Intelligence and Visa, reveals a widening performance gap between firms that have modernized their receivables and working capital infrastructure and those that continue to rely on manual, legacy processes.

This shift does not happen overnight. It requires a recalibration of incentives, a willingness to challenge assumptions and a commitment to continuous learning. But as the competitive landscape evolves, the cost of inaction becomes increasingly difficult to ignore.

Across major economies, the implications of these payment innovations are reshaping how businesses manage liquidity, reconcile transactions, and expand across borders.

In Europe and the U.K., open banking and real-time networks are driving innovation in embedded finance. In Asia, integrated ecosystems and cross-border connectivity are redefining the role of payments within digital platforms. In North America, APIs and data-driven approaches are enabling more sophisticated payment intelligence.

These differences reflect the underlying infrastructure and regulatory environments of each market. Many of them have resulted in an unavoidably complex and fragmented B2B payment landscape, one that complicates the day-to-day responsibilities of multinational treasury teams and CFOs.

Yet they also point to a broader convergence around a common theme: corporate payments as a system-level capability rather than a standalone function.

See also: Cross-Border Payments Hit a New Bottleneck at the Data Border 

From Rails to Systems

As corporate payment rails mature, differentiation is migrating upward. What matters now is not just how quickly or cheaply money moves, but how intelligently that movement is embedded within business processes. Payments are becoming programmable, contextual, and data-rich, serving as components of broader financial workflows rather than as isolated transactions.

“The office of the CFO is broadening its mandate,” Boost Payment Solutions Founder and CEO Dean M. Leavitt told PYMNTS last month, adding that while decisions about how companies pay and are paid “were traditionally a secondary issue for most CFOs,” that legacy hierarchy is now changing as finance leaders come to recognize the working capital implications of strategic B2B payment design.

This is evident in a growing emphasis across the marketplace on orchestration platforms. These systems allow companies to route payments dynamically across multiple rails, apply business rules in real time, and integrate payments directly into enterprise resource planning (ERP) systems. The result can be a more adaptive payments architecture that aligns with operational needs rather than constraining them.

Faster Payments in the U.K. and SEPA Instant in Europe have effectively commoditized speed. What differentiates providers now is how they enable businesses to use that speed, and the result is a payments ecosystem where the rails are largely interchangeable, but the systems built on top of them are highly differentiated.

In North America, where real-time payment adoption has been more gradual, innovation is being driven by API-based ecosystems and data-centric approaches. The emergence of new real-time networks is important, but the more significant trend is the proliferation of platforms that leverage APIs to integrate payments into enterprise workflows and the growing use of virtual cards.

“Virtual cards make it very easy to automate end to end,” Rene Stynen, senior vice president, EMEA, B2B Payments at Boost Payment Solutions, told PYMNTS in an earlier interview. “They enable straight-through processing and help build cost-sharing models between buyers and suppliers.”

That cost-sharing element is proving to be particularly important in Europe, Stynen added, where suppliers have historically resisted card acceptance due to processing fees.

Read also: CFOs Ditch AI Features to Fix Broken Payment Flows 

Redefining Liquidity, Working Capital

The shift from rails to systems is not merely a technological evolution; it is a redefinition of what payments mean within the enterprise. As the underlying infrastructure continues to improve, the real value will be created by those who can orchestrate it most effectively.

One of the most significant implications of this transition is the redefinition of liquidity management. In a world where payments are integrated into broader systems, liquidity becomes a dynamic resource that can be optimized in real time.

Businesses are increasingly using payment platforms to gain granular visibility into their cash positions, forecast future flows, and make decisions about when to pay or collect.

“All our research shows that access to capital and cash flow flexibility are the biggest friction points,” Mark Barnett, global head of small and medium enterprises at Mastercard, told PYMNTS this month, noting that true competition is not between payment mechanisms, but between constrained and unconstrained liquidity.

Nearly half of SMBs say they would pay for tools that let them adjust payment timing based on when they actually have money.

For all PYMNTS B2B coverage, subscribe to the daily B2B Newsletter.