RC: In terms of recent trends, could you provide an overview of how technological developments are helping to combat fraud and guide corporate compliance with cyber security procedures? Specifically, how does this relate to detecting malware and other online attacks more quickly?

Bradley: The concept of enterprise fraud management is changing. It still requires a focus on aggregating data across channels, and building holistic customer profiles, however with the application of behavioural analytics in real time, organisations can realise more accurate and faster decision making. In addition to this continuous monitoring, enterprise solutions are now optimising simulation and providing proactive forensic environments across a broader set of information. Through in-memory analytics, simulation that once took hours now takes seconds, allowing fraud analysts to test multiple methods. They can then deploy these new models much more quickly to better respond to flash fraud, and zero day threats that can manifest themselves in the online channel. Using storage technologies like Hadoop, more data can be stored and quickly surfaced within analytically driven discovery and visualisation interfaces, thus enhancing the investigative power and supporting identification of emerging threats. The analytics themselves are also advancing. Machine learning techniques are being utilised to continuously improve detection models based upon recent trends and indicators.

Jul-Sep 2015 Issue