TL;DR: Fraud teams face a control gap when real-time account-to-account fraud outruns single-bank defenses, according to Feedzai. Shared signals and outbound risk checks become critical as instant payment volumes climb toward 575 billion transactions by 2028.
Editorial analysis by NHI Mgmt Group, based on content published by Feedzai: “How Network Intelligence Is Reshaping Real-Time A2A Fraud Prevention”.
Key questions
Q: Why do social engineering attacks bypass many traditional fraud controls?
A: Social engineering succeeds because the real user is still on the real device, using valid credentials and a normal location.
Q: How should banks measure whether network intelligence is actually improving fraud prevention?
A: Look for reduced time-to-decision, earlier detection of suspicious beneficiaries, fewer false negatives across receiving banks and faster adaptation to new scam patterns.
Q: What breaks when fraud controls are built for batch review instead of real-time payments?
A: Batch-oriented controls lose effectiveness when payment decisions complete faster than human queues can react.
Practitioner guidance
- Prioritise outbound transaction scoring Score both sending and receiving activity before settlement, not only the transaction that originates inside your own environment.
- Build shared-signal workflows with network partners Exchange risk indicators, beneficiary patterns and mule signals in a privacy-preserving format so cross-institution patterns are visible in time to act.
- Add beneficiary risk to payment decisions Include high-risk beneficiary monitoring, mule detection and repeat-recipient analysis in the decision path for instant transfers.
Bottom line: Real-time A2A fraud exposes a control gap where instant settlement outpaces the ability of single-bank defenses to judge trust in time.
Glossary terms: Network Intelligence, Account-To-Account Fraud, Money Mule Account, Privacy-Preserving Signal Sharing
Explore further
View Full Forum → | NHI Foundation Course → | Our Services → | Read the full analysis →
Network intelligence is becoming a governance control, not just a fraud feature. The article shows that the core problem in real-time A2A fraud is not a shortage of signals but the lack of shared context at the moment of decision. That turns fraud prevention into a governance question about how quickly institutions can interpret risk across organisational boundaries. For practitioners, the control objective is collective visibility before money exits the account.
A few things that frame the scale:
- Nearly 60% of companies reported that fraud losses were still increasing in 2025.
A question worth separating out:
Q: What is the difference between isolated bank fraud controls and network intelligence?
A: Isolated controls evaluate risk inside one institution’s boundary, using only local data and history. Network intelligence adds cross-institution context, so a beneficiary that appears clean to one bank can still be flagged because of patterns seen elsewhere. The difference is not just scale, but the ability to detect coordinated fraud that crosses organisational silos.
👉 Read our full editorial: Real-time A2A fraud now depends on shared signal networks