CFOs have typically solved payments problems by buying more payments technology.
A new rail for faster settlement, another provider for cross-border payments. Fraud software, treasury platforms, AP automation, virtual card capabilities, and countless APIs connecting everything. Of course, it usually then took more software to reconcile what all that other software was doing.
Each decision made sense individually. But now that CFOs have arrived at the future they were building their payments and finance stacks for, many are finding instead that, taken together, these modernization initiatives that were deployed as point-solutions across the enterprise only created new problems.
Corporate payments infrastructure, broadly speaking, has become an expensive web of overlapping providers, workflows, controls and data, leaving finance teams paying a growing complexity tax just to move money efficiently.
Artificial intelligence could finally give CFOs a reason to start tearing some of it down.
See also: The Finance Stack’s Great Unbundling Has CFOs Asking What They Need to Own
The Payments Stack Has Become a Payments Mess
The biggest AI opportunity in payments may not be automating another task or bolting another copilot onto the finance stack. It may be moving intelligence above individual rails and providers, allowing companies to consolidate the sprawling infrastructure they spent years assembling.
“It’s not about abandoning legacy systems, but modernizing around them intelligently,” Garrett Baird, vice president of product, banking and FinTech at Paymentus, told PYMNTS in May.
Corporate payments infrastructure, after all, is rarely designed. It accumulates. That’s why the next payments upgrade could be subtraction. Companies can end up with multiple platforms performing adjacent functions, multiple sources of transaction data and multiple control environments governing different stages of the same payment.
Complexity itself becomes an operating expense. It is also a slippery one because the cost is scattered across software budgets, bank fees, headcount, implementation projects and operational risk.
Read also: The 4 New Working-Capital Levers CFOs Can Pull in Real Time
There is an irony here: Payments innovation largely worked. Businesses have more ways to move money, more providers competing for transactions and more tools for optimizing how money moves.
“We’ve seen a shift in moving away from the batch mindset,” Matthew Miller, managing director, treasury product executive at Bank of America, told PYMNTS in an interview published Thursday (Aug. 20). “It’s no longer nine-to-five. It’s now happening nights and weekends. The digitization of our environments is driving more to that single flow.”
But abundance creates another problem. Someone has to decide among all those options.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
Should a supplier invoice travel over ACH, instant payment, wire or card? Which bank account should fund it? Which provider offers the best economics for that currency? Should the company pay now or preserve cash? Does the counterparty require additional screening? Does the transaction fall outside normal behavior?
Today, those decisions are scattered across treasury policies, ERP configurations, rules engines, individual applications and humans. Artificial intelligence creates the possibility of putting more of that intelligence in one place.
Read more: The Fed Rate Hike Is Making Corporate Cash Too Expensive to Ignore
Payments Won. Now It’s Up to AI to Control Them.
Much of the finance conversation has centered on using AI to process invoices faster, automate reconciliation, investigate exceptions or assist treasury teams. But if CFOs simply attach AI to every existing system, they could repeat the same mistake that created today’s sprawling stack: solving individual problems while making the overall architecture even more complicated.
A different opportunity is to use AI as a forcing function for consolidation. An intelligent orchestration layer could potentially evaluate payment method, counterparty risk, liquidity, FX exposure, transaction costs and internal policy before recommending how money should move. It could prepare transactions, validate information, route approvals and escalate exceptions while maintaining an audit trail of the decisions behind each action.
If intelligence becomes centralized, CFOs can start asking whether every platform underneath it still needs its own analytics, workflow, decisioning and control capabilities.
Of course, the finance team’s risk appetite is ultimately what shapes progress in the Office of the CFO. The PYMNTS Intelligence report “The Cost of Caution: Why CFOs Put Growth Plans on Hold,” the latest installment of the 2026 Certainty Project, found in September that middle market CFOs set a higher bar for investing than retreating. More than half said they need a high level of certainty before committing to expansion, yet 91% said only a small or moderate decline in certainty could push their companies into a defensive posture.
There is, of course, a catch. CFOs cannot drop an AI agent into treasury on Monday and start canceling software contracts Tuesday. But if they don’t get started today, then there’s a better than fair chance they’ll be outpaced tomorrow.
For all PYMNTS B2B coverage, subscribe to the daily B2B newsletter.
Add as Preferred Source
