How Can Finance Teams Stop Payments Fraud Before Money Moves?

Learn how finance teams can reduce payments fraud by validating suppliers, modernizing payment methods and using AI as part of a layered defense.

Conversations around fraud often focus on detecting suspicious transactions. But as technology rapidly changes the landscape, companies are finding they should switch the focus to preventing bad actors from entering the payment process in the first place.

That's partly because the most treacherous payments fraud isn't always the most exotic. AI-generated deepfakes and synthetic identities get lots of attention, but finance teams are more likely to encounter more commonplace attempts. The fakery could be any number of things: an email that looks like it came from a trusted supplier, a request to change banking details or a payment instruction that appears routine until someone checks the data and verifies that it’s authentic. Whatever the scam, access made it possible.

That's why it’s no longer considered a win to identify shady transactions after they’re in the system. The objective now should be to make the payments environment harder to exploit. That means modernizing methods, validating suppliers and bank accounts before money moves, applying layered controls, being fully ISO 20022 compliant, and connecting verification to trusted supplier networks that can help confirm whether a payee relationship is legitimate before funds leave the business.

In Prevention First: Building a Smarter Defense Against Payments Fraud, a new PYMNTS Intelligence report produced in collaboration with Bottomline, we learn that half of respondents said AI-written vendor or supplier impersonation emails are the AI-enabled fraud tactic they encounter most often, while 88% see them at least a few times a year.

That’s from a 2026 survey of 150 U.S. treasury and finance executives at companies with at least $100 million in annual revenue. Digging into that research, we see how corporates are detecting and preventing fraud that continues to wreak havoc with payments.

The Threat Looks Familiar, but AI Raises the Stakes

Stopping fraud before it gets past countermeasures is easier said than done given the new tech. As stated, AI-written vendor or supplier impersonation emails are the top fraud tactic respondents encounter. Eighty-eight percent see them at least a few times a year, and 32% encounter AI-enabled business email compromise (BEC) at least monthly.

AI is improving the quality and scale of fraud attacks that finance teams know well. As the report puts it, “How a payment gets sent, who gets verified before it goes out and what technology watches over the transaction all shape how exposed an organization is.”

Paper checks remain a stubborn vulnerability. Fully 35% of firms reported experiencing check fraud, and nearly half (48%) said paper checks still account for at least 10% of outbound fraud losses, while 15% attributed more than one-quarter of those losses to checks. The implication is straightforward: modernization isn’t separate from fraud prevention. Ending reliance on risky throwback payment methods can limit and even remove exposure before another detection layer is needed.

Manual Controls Carry a Hidden Operating Cost

Strong controls work, but they can be expensive in time and attention. Eighty-nine percent of finance leaders called vendor pre-payment verification and onboarding a moderate to major burden. Seventy-eight percent said the same about reconciling payments.

That drain is an important yardstick, because prevention effectiveness is hard to measure by tallying losses that never happen. But manual verification consumes scarce finance resources. The report makes the point directly, saying, “Fraud prevention and operational efficiency are increasingly the same challenge.”

The better model is to preserve strong verification while automating more of it. Eighty-three percent of respondents use bank account validation to identify fraud, and 82% require more than one person to approve adding a vendor or changing bank details. Those controls create a trusted baseline against which activity can be checked and fraud detected.

That is also where network-based approaches can add something finance teams can't easily build alone: shared intelligence. A trusted supplier network can help validate vendor identity, banking credentials and payment behavior against a broader set of known relationships, rather than forcing each buyer to verify the same supplier in isolation.

At all times, a good CX must be maintained. Fraud prevention can't become a gauntlet for legitimate vendors. Stronger onboarding should reduce friction, not multiply requests for forms, emails, and follow-up calls. The more verification can happen through trusted data and reusable network credentials, the easier it becomes for suppliers to adopt secure processes without slowing down payments.

AI Works Best as a Layer, not by Itself

The survey reveals an unusually sharp divide in AI adoption. Fifty-seven percent of firms use no AI-powered fraud detection tools at present, while 42% use three or more. Just 1% sit in the middle.

Among adopters, however, the results are significant. Seventy-seven percent said AI fraud detection is more effective than the methods used before. More importantly, 59% of firms running three or more AI tools said they stop roughly 90% of fraud attempts before any loss occurs, compared with 32% of firms using no AI fraud tools.

Impressive as it is, the data doesn’t support replacing established controls with AI. Bank account validation is used by 83% of firms, manual review by 77%, and third-party fraud platforms by 65%, compared with 43% using AI or machine learning for fraud detection. That means AI is another layer in the line of defense, not as standalone fraud fighting technology. Just as important, AI works best when it has high-quality signals to work with, including verified vendors, trusted payment relationships, and network intelligence that helps distinguish true anomalies from routine supplier activity.

One respondent described that layered approach in practice: “An AI-written email impersonating a vendor requested updated banking information, but our account validation process flagged the change and prevented the payment from being sent.”

The Holdouts are not Standing Still

The biggest barrier to AI adoption is not necessarily skepticism about whether it works. Among non-users, 55% said manual reviews already handle fraud well enough. Meanwhile, 89% cited fear of delaying legitimate payments; 74% want human review of automation.

Current AI users have many of the same concerns. Ninety-two percent said fear of false positives made adoption harder, while 89% cited reluctance to automate without human oversight. Cost and legacy integration also remain obstacles.

Despite this, the direction of adoption is easy to see. Well over half (58%) of firms not currently using AI say they are implementing it now or will within 12 months.

Just 2% say they have no plans to adopt AI.

Prevention Becomes the Investment Priority

Among the most telling findings is where companies plan to spend. Spoiler alert: AI doesn’t top the list. Vendor onboarding verification ranks first, with 81% of firms saying they are likely to adopt or improve it over the next 12 months. Prepayment bank account validation follows at 75%, with AI-driven transaction monitoring at 73%.

That ordering captures the larger pattern. Finance leaders aren’t choosing between human controls, verification technology, payment modernization, trusted supplier networks and AI. They are building a stack in which each layer reinforces the others.

That's where trusted ecosystems and network-based verification become especially valuable. When payer, supplier, and bank account details can be validated against a shared network of known relationships and payment behavior, finance teams get an added layer of confidence before funds leave the business. It moves prevention beyond checking a single transaction and toward confirming that the entire payment relationship is secure. It also creates a better experience for suppliers, because verified information can be reused instead of re-collected every time a vendor is onboarded, updated, or paid.

Get the Report Prevention First: Building a Smarter Defense Against Payments Fraud