Most banks monitor fraud the way they are organized internally: a card team watching card risk, a UPI team watching UPI risk, a lending team watching lending risk. It's a reasonable structure for running a bank. It is not how fraud actually moves.
A single compromised identity can touch every one of these in sequence — a takeover on the mobile app, a fraudulent transfer via UPI, a card added to a wallet, funds routed to a beneficiary, cash recovered through a merchant transaction. Each individual step, viewed inside its own channel's monitoring system, can look almost unremarkable. The pattern is only visible when the full sequence is viewed together.
An Enterprise Problem Wearing a Channel Disguise
This is the uncomfortable part: channel-specific fraud systems aren't failing at their job. They are succeeding at a narrower job than the one fraud actually requires. Fraud has become an enterprise problem, and monitoring architecture built around products rather than customers will keep missing the journeys that matter most.
Fraud is increasingly an enterprise problem, not a channel problem. The fix isn't a better card system or a better UPI system — it's a fraud view that doesn't care which system a customer happened to touch first.