For most of the history of payments, verifying whether a payment provider charged correctly was a problem everyone had and no one solved. The leakage was real, the cause was understood, and the work required to catch it was simply too large to justify. In 2026, that equation is changing. The scale of the problem is better understood than ever, the manual status quo is visibly straining under more providers and more complexity, and for the first time the technology exists to verify charges at the level they actually occur. This is a landscape view of where PSP fee verification stands in 2026: the size of the problem, why it persisted, and what is shifting.
TL;DR
- Fee leakage, the gap between what providers should charge under contract and what they actually charge, typically runs 0.2 to 0.5 percent of payment volume on cross-border platforms.
- The backdrop is a large and growing market: cross-border B2B payment flows are projected to rise from roughly $39 trillion toward $56 trillion by 2030, and cross-border inefficiency costs, FX markups chief among them, run to over $100 billion a year.
- The status quo remains manual: spreadsheets and reconciliation tools that confirm records agree but do not verify charges against contracts.
- Integration debt compounds the problem, with the average payments operation spanning six to ten vendors, each in its own format.
- The shift in 2026 is from impossible to feasible: AI and modern infrastructure now make per-transaction verification economically viable, and verification is emerging as a distinct category.
- For CFOs and investors, the ability to verify provider economics is becoming a marker of financial control rather than a nice-to-have.
Short answer
In 2026, PSP fee verification is moving from an unsolved structural problem to an emerging category. The problem is well quantified: fee leakage typically runs 0.2 to 0.5 percent of payment volume on cross-border platforms, against a backdrop of cross-border B2B flows growing from roughly $39 trillion toward $56 trillion by 2030. The status quo is still manual, spreadsheets and reconciliation tools that confirm records agree without checking charges against contracts, made worse by the integration debt of running six to ten payment vendors. What is changing is feasibility: AI and modern infrastructure now make it economically viable to reconstruct the expected cost of every transaction and verify it against the contract, work that was previously impossible at scale. The result is that verification is becoming a recognized layer of fintech infrastructure and a marker of financial control.
The scale of the problem
The case for verification starts with the size of what goes uncaught, and the numbers are no longer vague.
On cross-border platforms, fee leakage, the difference between what providers should charge under contract and what they actually charge, typically runs 0.2 to 0.5 percent of payment volume. For a platform moving $500 million a year, that is $1 million to $2.5 million leaking annually through charges nobody verified. The largest contributor is usually FX, where the cost is buried in the rate rather than shown as a fee; the inefficiency costs of cross-border B2B payments, with FX markups a major component, have been estimated at over $100 billion globally in a single year.
This sits against a market that is large and growing. Cross-border B2B payment flows are projected to climb from roughly $39 trillion toward $56 trillion by the end of the decade. As volume grows, a leakage rate that stays constant as a percentage represents an ever-larger absolute sum, which means the problem is not stabilizing with scale; it is expanding with it. The size of the wound, in other words, is both significant and increasing.
Why the status quo persisted
If the problem is this large and this well understood, the obvious question is why it went unsolved for so long. The answer is that the tools finance teams had were built for adjacent problems, not for this one.
Spreadsheets were the default, and they break under real transaction volume, cannot reliably encode conditional contract logic, and leave no audit trail back to the source. Reconciliation software did better at matching, but reconciliation answers a different question: it confirms that two records agree, your ledger against the bank or the provider, not whether the charge was correct against the contract. A charge can reconcile perfectly, both records agreeing on the same figure, and still breach the contract, so even rigorous reconciliation left fee leakage untouched. The reconciliation software market itself was valued in the low billions in 2025, a sizeable industry that nonetheless was not built to verify charges against contracts.
Compounding all of this is integration debt. The average payments operation now spans six to ten vendors, each reporting in its own format with its own fee taxonomy. Checking charges across that sprawl by hand is not a task that scales, and so, at exactly the volumes where leakage became most expensive, verification became least feasible. The status quo persisted not because finance teams were careless but because doing the work properly was, until recently, beyond reach.
What is shifting in 2026
The defining change of 2026 is that per-transaction verification has crossed from impossible to feasible, and a category is forming around that shift.
The broader fintech context is a move from experimentation to production: technologies that were pilots a year or two ago, real-time payments, stablecoins, AI-driven operations, are entering production workflows. Within that, two changes matter for verification specifically. First, there is a growing expectation that financial records be continuous and captured at the source rather than reconstructed from provider statements after the fact, which is precisely the posture verification requires. Second, AI and modern infrastructure have made it viable to turn dense contracts into executable pricing logic, normalize data across many providers, and compute the expected cost of every transaction automatically. The work that used to be uneconomic, reconstructing expected charges transaction by transaction, is now something software can do continuously.
The consequence is that fee verification is crystallizing into a distinct category of infrastructure, separate from reconciliation and billing, defined by the question it answers: was this charge correct against the contract? It is early, the category is still being named and understood, but the trajectory in 2026 is clear.
Why it matters for CFOs and investors
The shift has consequences for two audiences in particular.
For CFOs, the context is margin pressure. Cross-border margins are thin, and a leakage rate of a fraction of a percent of volume is material against them. The ability to verify provider economics, to know rather than assume that charges are correct, is becoming part of running a disciplined finance function, both for the recovery it produces and for the control it represents.
For investors, verification is increasingly a signal of maturity. A fintech that can show it verifies its provider charges and bills its clients on evidenced, traceable economics is demonstrating that it controls its margins rather than estimating them. In diligence, that distinction matters: it separates a company that hopes its economics are sound from one that can prove it. As the category matures, the absence of verification may start to read as a gap rather than a norm.
Outlook
The direction of travel is toward verification becoming a standard layer of the fintech stack, as expected as reconciliation or a ledger. Three forces point that way. Verification is moving from periodic to continuous, so discrepancies are caught while they can still be recovered. The manual work around it is increasingly handled by AI agents operating on a deterministic core, with humans retaining judgment. And the payment stack keeps growing more complex, more providers, more rails including stablecoins, more conditional terms, which strengthens the case for treating provider economics as a system to be verified rather than an invoice to be trusted. None of this is finished in 2026. But the year marks the point where verification stopped being an impossible aspiration and became an emerging discipline, with the tooling, the data, and the recognition to support it. Platforms built specifically for this, such as Bluefyn, are part of that emergence. Bluefyn analyzes transaction and provider data; it never moves, holds, or custodies funds.
The bottom line
The state of PSP fee verification in 2026 is one of transition. The problem is large and well quantified, fee leakage of 0.2 to 0.5 percent of volume against cross-border B2B flows heading toward $56 trillion, and it has persisted because the available tools, spreadsheets and reconciliation software, were built to confirm records agree rather than to verify charges against contracts. What is changing is feasibility: AI and modern infrastructure now make transaction-level verification viable, and a distinct category is forming around it. For CFOs it is becoming a matter of margin control, and for investors a signal of financial maturity. The work is not done, but 2026 is the year verification moved from impossible to inevitable.
Frequently asked questions
What is PSP fee verification?
It is the practice of checking whether the fees a payment service provider charges are correct against the governing contract, transaction by transaction, by reconstructing what each transaction should have cost and comparing it to what was actually charged. It is distinct from reconciliation, which only confirms that records agree.
How big is the fee leakage problem?
On cross-border platforms, fee leakage typically runs 0.2 to 0.5 percent of payment volume, so a platform moving $500 million a year leaks roughly $1 million to $2.5 million annually. The largest contributor is usually FX cost hidden in the rate, and cross-border B2B inefficiency costs, FX chief among them, have been estimated at over $100 billion globally in a single year.
Why hasn't this problem been solved before?
Because the available tools were built for adjacent problems. Spreadsheets break at volume and leave no audit trail, and reconciliation software confirms that records agree rather than that charges are correct against contracts. Reconstructing the expected cost of every transaction across many providers was too much work to do by hand, so it was not done.
What changed in 2026?
Feasibility. AI and modern infrastructure now make it viable to turn contracts into executable pricing logic, normalize data across providers, and compute the expected cost of every transaction automatically. Work that used to be uneconomic is now something software can do continuously, which is why verification is emerging as a distinct category.
Why do investors care about fee verification?
Because it signals financial maturity. A fintech that verifies its provider charges and bills on evidenced, traceable economics demonstrates that it controls its margins rather than estimating them. In diligence, that separates a company that hopes its economics are sound from one that can prove it, which is increasingly a meaningful distinction.
Is reconciliation software enough to catch fee leakage?
No. Reconciliation confirms that two records agree, such as your ledger and the provider's. It does not check whether the charge was correct against the contract. A charge can reconcile perfectly, with both records agreeing on the same figure, and still be an overcharge, so reconciliation leaves fee leakage uncaught. Verification against the contract is what catches it.
