Device fingerprinting
Device fingerprinting combines attributes such as browser, operating system, screen settings and software configuration into an identifier that recognises a returning device. It is probabilistic, so it supports a judgement about familiarity and does not prove who is using the device.
In more detail
The idea is that a customer's device is more consistent than their location. A login from a device the account has used before is lower risk; the first appearance of a device, especially alongside a credential change, is a signal worth weighting.
Fingerprints are not identity. They can change when software updates, can collide between identical devices, and can be spoofed by an attacker who emulates a known configuration. Privacy rules also constrain how device data may be collected and kept. So a fingerprint is best treated as one weak signal among several, not as a decisive test.
A system receives the device attributes with each event, matches them against devices previously seen on the account, and adds to the risk score when the device is new.
Example
A customer has used two devices for a year. A login arrives from a third, which has never been seen on the account, and a new payee is added within minutes. The unfamiliar device raises the score, though it would not be enough on its own.
Related capability
Signals evaluated by transaction monitoring
Related terms
Discuss your transaction monitoring requirements
Tell us which transaction flows you need to monitor, which systems Akoben would integrate with, and how your fraud team works today.
Request a demo