Real-time fraud detection, part 1
A recent IBV study indicated that "most financial institutions can't detect fraud until after the money moves." With technological advancement creating so many more payment options are we opening ourselves to more fraud? Can we detect and counter fraud in real-time? Here are some of the questions IBM's David Dixon and and experts Cherian Abraham and Dave Birch discussed in the chat:
- Fraud has always been a top-of-mind concern for financial institutions. What are some of the fastest-growing fraud threats impacting financial service providers?
- In the past, fraud detection involved relying on customers to phone the call center to dispute a charge on their credit card bills. Fraud was fairly simple, opportunistic and individualized, such as when a lost credit card was picked up in the mall parking lot and used for a shopping spree. Now, 80 percent of consumer fraud is estimated to be perpetrated by organized criminal gangs using multiple product channels, multiple locations, multiple players and a very short—sometimes only within hours—campaign window for execution. How can financial institutions guard against these more sophisticated attacks?
- Real-time detection—the ability to interdict a fraudulent transaction before it is settled—is a technology that is currently available. How does it work and how can this technology change the game for fraud detection?
- Can we use fraud patterns in historical data to help minimize fraud? How?
- The most successful institutions employ individuals who have both deep analytical skills and seasoned expertise in countering fraud. How can a company that currently reports a cycle time of four weeks or more just to discover the pattern improve on these adjustments? Is there a skills gap for fraud experts with analytical skills?
- Many companies are actively centralizing operations to eliminate silo problems in handling complex issues. Would this approach help an organization better handle fraud issues?
- In fraud prevention efforts, the inclusion of additional relevant information in the analysis can translate into improved detection rates, minimized false positives and reduced operating costs for alert management and investigations. Are financial institutions employing the technology necessary to analyze data from multiple data sets—internal, external and unstructured—to detect fraud?
- As fraudsters become even more savvy, how can financial services organizations stay abreast or even get ahead?