Fraud detection systems

Fraud Detection is a technological line of defense for banks against fraudulent transactions, compromised devices, counterfeit documents, insider activity and other forms of abuse.  The systems use behavioral analytics, statistical and ML models, as well as pre-defined rules to identify anomalies in payments, accounting actions, authorization and communications.

Key functions:

• Real-time transaction monitoring (on-line fraud scoring);
• Behavioral analysis of clients (device fingerprinting, geo-IP, speed, rhythm, patterns);
• Detection of suspicious actions in online and mobile banking, ATMs, POS terminals;
• Identification of fraud by cards, with applications, documents, personnel (internal fraud);
• Integration with anti-fraud hacks and behavioral biometrics;
• Creation of models based on Machine Learning and scenarios (rule-based / hybrid);
• Generating alerts, initiating blocking, black and grey lists, integrating with AML and CRM.

Who works with the system inside the bank:

• Financial monitoring and security service
• Operational and client unit (incident handling)
• Digital banking / e-commerce and card teams
• IT, cybersecurity and risk systems architects
• Investigation and internal audit department

System owners:

• Chief Security Officer (CSO) or Head of Fraud Risk
• CRO (in terms of integration with the general risk model)
• CIO/CTO — technological support, integration, SLA