Mobile application behavior detection: the cheap way to catch fraud
After RSA's acquisition of SilverTail, things are heating up in mobile application level behavioral detection.
We see fraud management vendors increasingly looking at mobile application behaviors (beyond web fraud management and device fingerprinting) to build out a normal and abnormal behavior profile for the network traffic signatures coming out of the application (similarly to how SilverTail/RSA looks at web traffic signatures). Note that this is clearly a grey area that falls between what device fingerprinting vendors (iovation, 41st Parameter, BlueCava, ThreatMetrix), or risk-based authentication (RBA) vendors (RSA, Entrust, CA/Arcot, etc.) or what traditional back-end, cross-channel transaction monitoring vendors (Actimize, ACI, Detica, SAS, etc.) have been doing. Although device fingerprinting and RBA vendors have long been providing SDKs and APIs for developers to include in their mobile applications, understanding mobile application network traffic and building good and bad behavioral models is becoming something people are increasingly interested in.
Mobile application behavior detection has the benefits of not having to open up application code, not having to define too many security policies or rules. Because of this, mobile application behavior detection and network traffic signature profiling is something we expect to see a lot of vendor interest in the next 9-12 months.