Network view boosts fraud detection rates

The challenge posed by payments fraud in Europe continues to grow. In 2024, the total value of fraudulent credit transfers sent by payment service providers across the EU and EEA reached €2.5 billion.[i] As these figures capture only cases that are identified and formally reported, the true scale of the problem is likely to be considerably higher.

While SCT Inst volumes almost doubled between 2022 and 2024 alone, rising by 98%, the number of fraudulent instant payment transactions increased by 175%, indicating that SCT Inst fraud grew significantly faster than overall usage. This trend underlines the extent to which fraudsters favour instant payment rails.

As new vulnerabilities and fraud patterns emerge, how can PSPs move beyond institution-level controls to identify fraud more accurately and in real time? The answer could lie in network-level transaction monitoring.

The changing nature of fraud

The growth in authorised push payment (APP) fraud has fundamentally changed the challenge facing payment service providers. As one panelist at EBAday 2026 pointed out, more than 90% of fraud victims are now manipulated into initiating and authorising payments themselves, placing greater emphasis on institutions’ ability to assess whether the customer’s decision is genuine, alongside verifying their credentials and payment instructions.

The challenge is being amplified by two parallel developments: artificial intelligence is enabling convincing impersonation attacks, while the continued growth of instant payments leaves institutions with only milliseconds to assess risk before funds become irrevocable. In many cases, the payment is merely the final stage of a scam that began on social media, a messaging platform or another digital channel. Fraud prevention is therefore a broader ecosystem challenge, rather than one that payment service providers (PSPs) can solve alone.

Against this backdrop, Verification of Payee (VOP), introduced across the EU in October 2025, represents an important but necessarily limited safeguard. Designed to prevent misdirected payments, VOP confirms that the beneficiary name matches the account details provided, but it cannot determine whether the beneficiary is legitimate or whether the customer has been manipulated into making the payment. For example, a merchant may legitimately own an IBAN while still operating a scam.

VOP’s broader significance lies elsewhere. By encouraging PSPs to integrate APIs and consume shared network intelligence in real time, VOP has helped establish the foundations for a more sophisticated approach to fraud prevention.

Real-time transaction monitoring

One example of this next generation of fraud prevention capabilities is EBA CLEARING’s Fraud Pattern and Anomaly Detection (FPAD) functionality. Available to STEP2 and RT1 participants, FPAD complements banks’ existing fraud controls by providing account assessments, transaction assessments and investigation capabilities based on network-wide intelligence drawn from billions of transactions. Network risk indicators are returned in milliseconds, allowing institutions to incorporate external intelligence alongside their own behavioural analytics, device intelligence and machine learning models.

The value of this broader perspective is already evident, as testified in a panel discussion among FPAD users at EBAday 2026. One of the participating PSPs has integrated pan-European FPAD risk indicators into its existing fraud monitoring framework, combining its own analytics with network-wide intelligence. Using only FPAD’s core indicators, this institution has more than doubled its fraud detection rate to over 60%, demonstrating how network intelligence can strengthen, rather than replace, existing controls by revealing fraud patterns that would otherwise remain invisible. This is confirmed by the overall feedback EBA CLEARING has been receiving from FPAD users: they have seen their fraud losses reduced by 30% to 60%, depending on the structure of their customer portfolio. To further broaden the value delivered by FPAD, the Company continues to develop the functionality, with several major enhancements now underway:

1. Expanding to direct debits: FPAD is being extended to direct debit transactions, which are increasingly a target of fraudsters. While an individual fraudulent direct debit may still need to be rejected or recalled, the real value of FPAD lies in its network-wide perspective. As direct debit transactions are submitted the day before settlement, participants can receive early warnings of emerging fraud patterns, enabling them to take preventive action before the payments are processed.
2. Mule account detection across institutions and borders: Money mules form the backbone of all fraudulent activities. Although it is not impossible to follow the money manually from one mule account to the next, across institutions, PSPs are now able to implement a more systematic and real-time approach to detect mule account networks based on FPAD data connecting these accounts to fraudulent transactions. At the request of its users, EBA CLEARING is also exploring the use of AI (neural networks) to further increase the efficiency of this approach.
3. Learning from participant feedback: FPAD’s passive network intelligence is now being complemented by active feedback from participants. Institutions consuming FPAD signals can share investigation outcomes and insights from their own fraud monitoring systems, creating a continuous feedback loop that refines the models over time. The result is more accurate risk indicators, better filtering of false positives and sophisticated fraud alerts, which can also be enriched by adhering to the EBA Fraud Taxonomy.

Dovetailing with the PSR

The direction of travel is clear. Effective fraud prevention will depend less on standalone controls and more on combining multiple layers of intelligence – from PSPs’ own behavioural and device analytics to insights drawn from across the wider payments network – to provide a far richer and more nuanced picture of payment risk than any individual control can achieve in isolation. This can help institutions identify more genuine fraud while also distinguishing legitimate transactions from false positives, which, in turn, can help to reduce unnecessary friction for customers.

This approach also aligns closely with the direction of European regulation. Under the Payment Services Regulation (PSR), PSPs will be expected to monitor incoming and outgoing transactions in real time and participate in information-sharing arrangements. FPAD already supports many of these capabilities, providing participants with a practical solution that is operational today and continues to evolve alongside regulatory requirements.

For more information, please read EBA CLEARING’s recent white paper: Real-time transaction monitoring: The future of fraud fighting.

__________

[i] Source: 2025 Report on Payment Fraud by EBA & ECB, issued in December 2025