FedNow Service: Definition and Use in Compliance

What is FedNow Service?

The FedNow Service is the Federal Reserve's instant payment rail. It settles transfers between U.S. banks and credit unions in seconds, every hour of every day, with funds available to the recipient immediately and the payment final on the spot.

It went live on July 20, 2023. Before that, most U.S. interbank transfers ran on ACH, which clears in batches and can take a day or more, or on wire systems that close on weekends and holidays. FedNow closed that gap by offering a always-on rail operated directly by the central bank.

Two design choices shape everything compliance teams do with it. First, settlement is gross and immediate: each payment clears on its own, in central bank money, the moment it's sent. Second, it's irreversible. There's no chargeback, no return window, no batch reconciliation step where a bank can pull a transfer back.

A practical example: a small business in Ohio pays a supplier in Texas at 9 p.m. on a Saturday. The supplier's bank credits the account within seconds, and the supplier can spend the money immediately. On ACH, that payment would sit until the next business day.

FedNow uses the ISO 20022 message format and supports request-for-payment, where a biller sends a payment request the customer approves. The Fed set the default transaction limit at $500,000 at launch and raised it to $1 million in 2025, though each bank can set lower caps. According to the Federal Reserve, participation is open to any depository institution with a Federal Reserve master account.

How is FedNow Service used in practice?

Banks use FedNow to offer customers instant payments for payroll, supplier invoices, insurance payouts, loan disbursements, and account-to-account transfers. The appeal is speed and certainty: the recipient knows the money has arrived and can't be clawed back.

For compliance and fraud teams, that same certainty is the problem. With ACH, a bank has hours to review a batch and return a suspect entry. With FedNow, the decision window is the few hundred milliseconds before the payment is authorized. Everything moves to real-time scoring.

In practice, teams build layered controls at the send moment. They set send limits per customer segment, often well below the network cap. They hold transfers to brand-new payees for a step-up check. They run behavioral analytics that compare a transaction against the customer's normal pattern, so a $40,000 transfer from someone who usually sends $200 triggers friction.

Consider a credit union that joins FedNow and sees a spike in small accounts receiving instant credits and immediately forwarding them out. That's a classic mule pattern. The team uses network analysis to map the receiving accounts, freezes the cluster, and files reports.

The metric that matters shifts too. On reversible rails, recovery rate is a useful number. On FedNow, the question is detection-before-authorization rate, because once a payment clears, the money is usually gone. Teams measure how many bad payments they stop, not how many they recover.

FedNow Service in regulatory context

FedNow doesn't change a bank's underlying obligations. The Bank Secrecy Act still applies. Banks must run customer due diligence, monitor for suspicious activity, file SARs, and screen against sanctions lists no matter how fast the rail moves.

What changes is timing. The Office of Foreign Assets Control expects sanctions screening on payments, and FedNow gives no batch window to screen after the fact. Screening against the SDN list has to happen inline, before authorization, and return a clean or hit decision in real time. A slow screening engine isn't a performance problem on an instant rail; it's a compliance failure or a customer experience failure, and often both.

FinCEN has flagged the fraud risk directly. In 2023 it issued an alert on the rise in mail-theft-related check fraud and, separately, has warned about the speed at which instant payments let criminals move and layer funds. The Federal Reserve also published fraud-mitigation guidance and built network-level tools, including a negative-list feature banks can use to block known bad accounts.

A concrete regulatory pressure point: a bank that onboards a customer with weak KYC and then lets that account send instant payments has compressed the window to catch a problem from days to seconds. Examiners increasingly ask how a bank's monitoring keeps pace with the rails it has enabled. The Federal Reserve's FedNow fraud resources set out expectations for participants.

Common challenges and how to address them

The hardest challenge is latency. A monitoring system that returns a decision in two seconds is fine for ACH and useless for FedNow. Banks address this by moving rule evaluation and model scoring inline, often pre-computing customer risk profiles so the real-time decision is a fast lookup plus a few live signals rather than a full recalculation.

The second challenge is fraud, especially authorized push payment fraud. Because the customer authorizes the payment themselves, traditional account-takeover defenses miss it. The victim's credentials are fine; they've simply been deceived. Banks counter this with payee-confirmation checks, transaction-pattern warnings shown to the customer mid-flow, and cooling-off holds on high-value first-time transfers.

Third is the mule problem. Stolen funds get scattered across receiving accounts within minutes. Effective teams use graph analytics to spot accounts that receive and immediately forward funds, and they share signals across the FedNow negative list where permitted.

A fourth issue is alert volume. Real-time scoring on a 24/7 rail generates alerts at all hours, and a tired team makes worse decisions. Reducing false positives through better feature engineering and tighter threshold tuning is how banks keep the queue manageable without missing real fraud.

The honest tradeoff: tighter controls add friction and can hold legitimate payments, which undermines the speed customers signed up for. Every bank running FedNow lives somewhere on that curve, and the right point depends on its risk appetite and customer base.

Related terms and concepts

FedNow sits alongside other instant and real-time payment systems. The private-sector RTP network launched in 2017 and competes directly with it in the U.S. Internationally, the UK's Faster Payments Service and India's Unified Payments Interface are comparable rails, each with its own fraud history that U.S. banks study for lessons.

On the compliance side, FedNow connects to nearly every core financial-crime concept. Transaction monitoring has to run in real time. Sanctions screening has to clear inline. Suspicious payments still end in a SAR filed with FinCEN, and large cash-funded transfers may still trigger a Currency Transaction Report.

The fraud typologies most associated with instant rails include account takeover, synthetic identity fraud used to open mule accounts, and mule networks that disperse funds.

Teams building real-time defenses often pair FedNow controls with AI-powered fraud detection and broader transaction monitoring rule tuning to keep decision latency low without letting fraud through. Understanding FedNow well means understanding the rest of the instant-payments and financial-crime stack it plugs into.