Fraud is becoming more pervasive, more adaptive and more difficult to contain, and financial institutions are responding by reworking the way detection, prevention and customer protection are embedded across payment systems.
The Federal Reserve Financial Services Risk Officer Survey, released Tuesday (April 28), based on responses from more than 400 institutions, presents a systemwide view of that pressure. The central bank has found that fraud is increasing across debit, check, ACH and wire channels, driven by evolving criminal tactics, greater digital exposure and sustained scam activity that exploits both customers and institutions.
Debit card fraud remains nearly universal, while check fraud continues a multiyear resurgence. At the same time, institutions report rising exposure to account holder scams, business email compromise and unauthorized debits within ACH activity, along with persistent wire fraud tied to account takeover and mule network.
The survey emphasizes that no major fraud category is declining. As a result of that upward trajectory, institutions face a pattern of cumulative pressure in which multiple vectors, including impersonation and credential compromise, are advancing at the same time.
Operationally, the strain is compounded by detection challenges. Many institutions continue to identify fraudulent activity only after losses occur, particularly in the case of mule accounts, where funds are often depleted before intervention is possible.
PYMNTS Underscores Unauthorized Access
PYMNTS Intelligence findings reinforce and extend the Federal Reserve’s observations. “The 2025 State of Fraud and Financial Crime” in the United States report, in joint effort with Block, finds that unauthorized-party fraud now accounts for 71% of incidents and losses, reflecting a shift toward account takeover and credential misuse.
Advertisement: Scroll to Continue
This evolution signals a change in how fraud operators, well, operate. Rather than relying primarily on manipulating legitimate users into initiating transactions, attackers are gaining access to accounts and initiating activity directly. That distinction alters the defensive posture required of financial institutions, as prevention must occur earlier in the interaction and often before a transaction is initiated.
The consequences extend beyond financial loss. PYMNTS Intelligence reports that half of institutions experience negative effects on customer loyalty, while significant shares cite reputational damage and lost business opportunities.
The Federal Reserve survey dovetails with these findings by highlighting the rise in impersonation-based scams and credential-driven attacks across payment channels. Both datasets point to identity and access as the central battleground.
Technology as Defense
Faced with these pressures, financial institutions are increasing their reliance on technology to identify and mitigate fraud. PYMNTS Intelligence reports that 68% of institutions increased fraud detection spending, reflecting a shift toward AI-driven and behavioral analytics-based approaches.
These tools are designed to operate continuously, analyzing transaction patterns, user behavior and contextual signals to identify anomalies that may indicate account takeover or fraudulent intent. The objective is to move from reactive identification toward earlier detection within the transaction life cycle.
The Federal Reserve survey identifies similar approaches. Institutions cite real-time monitoring alerts, enhanced authentication controls and machine learning models as key components of their fraud mitigation strategies.
Additional measures include biometric verification, transaction monitoring and customer education aimed at reducing susceptibility to social engineering attacks.
At the same time, the survey underscores persistent limitations. Detection processes remain heavily manual in many institutions, system integration challenges delay response times and information sharing across institutions is often constrained. These factors contribute to delayed identification and reduced recovery rates.
From Fragmented Defenses to Integrated Systems
The combined findings suggest that the central issue is not simply the presence of fraud but the fragmentation of defenses. Effective mitigation requires coordination across systems, channels and institutions, as well as alignment between detection tools and operational processes.
Institutions that rely on isolated controls or post-transaction review are less likely to contain losses in an environment where fraud occurs at speed and across multiple touchpoints. By contrast, those that integrate analytics, monitoring and response into a continuous process are better positioned to identify and address threats before they escalate.
