ThreatFabric Says Behavioral Analytics Raised Banks’ Fraud Detection Rates

Summary

ThreatFabric reports that device intelligence and behavioral analytics improved fraud detection rates by 20 or more percentage points in several bank deployments, with alert volumes remaining largely stable in one example.

Key points

• ThreatFabric says transaction-focused risk engines typically detect 30%–50% of attempted fraud value, particularly struggling with scams victims authorize themselves.
• In one bank example, reported fraud visibility rose from 50% to 73% with device intelligence, then to 91% after adding behavioral risk analytics.
• Other deployments reported detection increasing from 65% to 90%, and from 61% to 85%.
• Signals cited include unusual navigation, deviations from customer behavior, typing and mouse patterns, device compromise, remote-access activity, and malware indicators.
• One deployment maintained about 30 alerts per day, with alert rates below 1% when combined with transaction data.
• The article attributes the largest gains to socially engineered fraud, including family, social media, impersonation, romance, and investment scams.

Vendors

Products