Fraud rule
A fraud rule is a condition on transaction data paired with an action, such as raising an alert or adding to a score, that is applied whenever a transaction matches the condition. Rules are explicit, so an analyst can always say why a particular transaction was flagged.
In more detail
Rules range from a single test, such as an amount above a limit, to composite conditions that need several facts to hold together. Composite rules are usually what separates a genuine pattern from noise: a large payment is ordinary, a large payment to a payee added an hour ago from an unfamiliar device is not.
Explainability is the lasting strength of rules. When a regulator, a customer or an internal reviewer asks why something was flagged, the answer is the rule. The weakness is upkeep: fraudsters adapt, so rules age, and a rule set that is never reviewed drifts into either missing fraud or flagging everything.
The engine evaluates each incoming event against every active rule, and a rule that matches triggers its action, such as raising an alert or adding to the score.
Example
A rule reads: if a payee was added less than an hour ago and the amount exceeds the account's usual maximum and the device has not been seen before, add a high score and raise an alert for review.
Related capability
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