Anti-fraud

Anti-fraud is the set of controls a lender, BNPL provider or payments company uses to detect and stop fraud in applications and payments. Checks and rules, fraud scores and analyst review help assess uncertain cases, and confirmed fraud can inform later checks.

Example

For illustration, a lender sees two applications under different names from the same device. Its rule sends the shared-device match for analyst review. If fraud is confirmed, that device can be added to a block list for later applications.

The checks used on applications and payments

Identity checks compare the applicant’s name with their date of birth, address and ID number, then compare those details with an identity document or a register. A device check fingerprints the phone or browser. The same device appearing on applications under different names stands out, as does an emulator.

IP checks look for proxy or VPN use. They can also identify hosting addresses and compare the IP location with the stated address. Email and phone checks consider how old an address or number is, and whether an email domain is disposable. Velocity checks count how often a detail such as a device, email or card appears within a time window. Card checks compare card details with the IP location and stated address.

From checks to a fraud decision

The lender’s rules translate check results into a decision under its policy. A fraud score ranks uncertain cases, which analysts can work through in a review queue.

Confirmed fraud can be added to block lists. A later application with the same device or email can match the list, feeding the confirmed case back into future checks.

Questions we get asked

In fraud decisioning, a confirmed-fraud list matches each application detail separately, then rules use the number of matches to decide.

The fraud rules engine article explains how confirmed fraud feeds block lists and how rule changes are backtested against past decisions.