Home / Wiki / R / Risk Engine

Risk Engine

Technology and VendorsUpdated 16 Nov 2026

A risk engine is software that assesses the risk of each transaction, login or onboarding application in real time and returns a decision such as approve, decline, send for review or ask for further authentication. Payment providers, merchants, banks and lenders use it mainly to stop fraud, and some systems also judge credit or compliance risk. The engine combines a rules engine, which applies fixed conditions such as velocity checks or country blocks, with statistical or machine learning fraud models that produce a risk score.

Inputs include the amount, card and account history, IP address, device fingerprinting data, behavioural signals and lists of known bad actors. Thresholds set the balance between stopping fraud and turning away genuine customers through false declines. A medium score might trigger step-up authentication, such as a 3-D Secure challenge, rather than an outright decline. Analysts review flagged cases, and their findings feed back into the rules and models.

Last updated 16 Nov 2026