Why businesses are getting fraud prevention wrong

For digital businesses facing AI-assisted fraud, tighter controls can look like the obvious answer. The catch is that aggressive blocking can also push away legitimate customers, which is a much uglier problem for PSPs and merchants than a neat-looking fraud dashboard.

1. Fraudsters are using AI to pressure the KYC (Know Your Customer) layer in particular. Volodymyr Todurov, chief executive of Frogo, says real-time document generation and deepfake face-swapping are becoming “remarkably sophisticated,” which makes established verification processes easier to attack at speed.
2. The commercial downside of heavy-handed controls is straightforward: suspicious payments may be stopped, but legitimate payments can be rejected too. That means delayed withdrawals, frustrated customers, and business being pushed elsewhere, even while the company can point to a lower recorded fraud rate.
3. Frogo, which combines device intelligence, automated risk scoring and real-time alerts, argues that fraud prevention should be judged as a revenue optimisation engine rather than a pure loss-prevention tool. Todurov breaks that into three measures: direct loss prevention, operational efficiency, and retention and lifetime value.
4. On the operations side, the question is not just how much fraud is blocked, but how much manual work it takes to get there. Todurov points to monthly active users supported per analyst, plus the time needed to complete investigations, as a better way to judge whether automation is actually helping the risk team scale.
5. The third measure is the one many businesses ignore until it is too late: what happens to legitimate customers. Todurov argues for “automated withdrawals built on the presumption of innocence,” and warns that a “guilty until proven innocent” approach can destroy customer experience and, in practice, block high-revenue players without anyone noticing in the fraud report.