Online Banking Fraud Detection: The Risk Engine

The FraudMAP® Risk Engine detects new and emerging online banking fraud schemes using predictive models of individual online customer behavior to monitor transactions and more. It differentiates normal online customer behavior from suspicious activity – including account breaches that leave no fraudulent transaction to detect.  The Risk Engine’s proprietary technology uses known information about fraud threats when available, but is not dependent on pre-defined fraud rules or transaction monitoring.  It generates a rich context for fraud investigations with the Risk Application.  To ease integration with the customer’s online channel, the Risk Engine features both a real-time API and batch controller for wider integration and deployment options.

Dynamic Account Modeling™ Understands Each Customer

The core technology of the Risk Engine is Dynamic Account Modeling™, which uses predictive models of each individual online account holder’s behavior session by session.  Because the Risk Engine is not dependent on pre-defined fraud rules and automatically detects anomalous behavior, new threats are detected as they occur. Because its activity modeling spans the gap from logins to transactions, the Risk Engine can reveal the previously undetected online steps of multi-channel fraud schemes before any loss occurs.

The Risk Engine easily handles real world situations such as changing customer and fraudster behavior, the use of proxies, corporate firewalls, dynamic IP addresses, and upgrades to customer hardware and software. Its activity modeling is an extensible framework that monitors a range of routine in-session activities and uniquely weighs inherent risk by a number of parameters, including size of potential loss.  The Risk Engine’s advanced statistical models are based on probabilities that dynamically adjust to individual account holder’s behavior across their entire session history, recognizing that every customer behaves differently and what might be unusual activity for one customer may be normal activity for another.  Measures of risk are driven by what the system expects the customer to do in a particular session and the “cost factor” of the individual activity.