It\u2019s never been harder for businesses to manage the integrity of their payment operations. Fraud schemes are evolving as bad actors leverage artificial intelligence, making threats more difficult to detect in real time. At the same time, faster payments leave businesses with less room for error and are more reliant on automation\u2014one of the very forces fueling AI-driven fraud. Only 8% of surveyed companies report no problems on this front.<\/p>\n
As a result, pressure is building in the payments industry for stronger \u201cupstream\u201d safeguards that can detect identity and bank account risks early in the transaction process. The urgency arises amid warnings from regulators, including the Financial Crimes Enforcement Network, about the growing use of deepfakes\u00a0and synthetic identities to bypass financial institutions\u2019 verification systems.<\/p>\n
It\u2019s why the National Automated Clearing House Association, which governs, develops, and administers the rules for the ACH Network, put new rules<\/a> into effect in March that require all non-consumer ACH participants to strengthen their monitoring of fraudulent transactions. The rules require two new standardized payment labels\u2014payroll or purchase\u2014on ACH transactions to help banks spot anomalies. Aimed at combating \u201ccredit-push\u201d fraud, where criminals trick businesses into initiating seemingly legitimate ACH payments, participants must establish risk-based monitoring processes across all transaction types, with the rule phased in on March 20, 2026, for large originators and on June 22, 2026, for everyone else.<\/p>\n All of this puts greater focus on the steps that verify the process for transferring money to a business or customer for goods or services delivered. The AR function is a delicate balancing act involving speed, customer experience, and risk management. Most companies still rely mainly on detecting fraud or nonclearance after settlement, when losses have already occurred, and recovery options are limited and uncertain. As a result, operational friction is rising, and the cost of missed or delayed detection is becoming more visible.<\/p>\n PYMNTS Intelligence\u2019s latest Certainty Project<\/strong><\/b> research shows that leading middle-market firms with at least $100 million in annual revenue are shifting their focus much earlier in the transaction process. By verifying bank accounts and identities earlier\u2014even before payments are initiated or authorized\u2014they can block AR integrity risks rather than absorb losses and attempt recovery later.<\/p>\n These are just some of the findings detailed in \u201cEarly Detection: Why Top Performing Firms Focus on Fraud Before it Starts<\/strong><\/b>, sponsored by Plaid<\/strong><\/b>. This edition examines the challenges surrounding fraud prevention and identity verification in AR processes. It draws on insights from a survey of 60 heads of payments at U.S.-based middle-market companies with annual revenues between $100 million and $1 billion. The study was conducted from March 18-30, 2026.<\/p>\n [branded_divider]<\/p>\n When a business sends out goods or services and waits to receive payment, there\u2019s a window of time during which things can go wrong. More than half of the businesses surveyed said they typically discover a problem only after the payment has already settled, which can mean the money is gone and getting it back is an uphill battle.<\/p>\n What separates the companies that catch problems early from those that don\u2019t is how well they verify who they\u2019re dealing with. The firms with better early-warning systems aren\u2019t just doing more legwork\u2014they\u2019re also using technology that connects directly to a customer\u2019s bank to confirm, in real time, that the account exists and belongs to the right person. This kind of check, combined with automated fraud scoring, gives companies a chance to pause or redirect a risky payment before it\u2019s too late. Firms that rely mainly on manual document reviews and phone confirmations, while common, tend to be the ones still learning about problems after the fact.<\/p>\n For nearly six in 10 (57%) firms surveyed, fraud or nonclearance is typically detected after respondents treat the payment as \u201cdone\u201d and only learn of the failure later via returns, disputes, or outreach from the receiving bank. Only three in 10 usually find these issues early, either before the transaction is initiated (17%) or during authorization (13%). High-uncertainty firms are slightly more likely than others to detect fraud or nonclearance early, but late detection is the norm across all three certainty tiers.1<\/a><\/sup> Results are also very similar across revenue brackets.<\/p>\n\n
Stuck in Reactive Mode<\/h2>\n
For most firms, AR integrity remains a backward-looking process: Nearly six in 10 businesses mainly detect fraud or nonclearance after settlement.<\/h3>\n