As money laundering schemes are getting more and more sophisticated, the detection and mitigation strategy has to get smarter and faster. Looking ahead to 2018, I predict a rising trend for organizations to do four things better than today in the realm of financial crime.
1. Take a More Holistic ApproachToday fraud prevention, cybersecurity and anti-money laundering (AML) are still siloed at many financial institutions. A holistic approach would have one customer risk classification used that would consider any available information, whether that comes from a fraud system, the CISO’s system or a compliance system.
2. Use Artificial Intelligence to Find New PatternsToday’s detection seems to be too static, with fairly basic analytics and an over-reliance on rules. For 2018 we see an increasing trend in adding artificial intelligence and more powerful algorithms to that capability.
The objective isn’t just to jump on the AI bandwagon — it’s to further decrease workload and increase productivity at compliance departments. We do not see a replacement of the traditional approach in the short term, but an enhancement with AI that takes detection to the next level. This will help prioritize the existing alerts and generate new alerts, as we “find the unknown patterns”.
3. Automate the Search for Tax EvasionCompliance is related to tax evasion in many cases. Recently we have seen the Panama Papers and Paradise Papers, and we will see more Papers leaks next year. Compliance organizations will be exploring the holistic approach to detection, and adopting tools that include any of this information in an automated manner.
4. Accelerate to Real-Time DetectionLast but not least: Criminal schemes are not only getting more sophisticated, and they will get faster as well. So must detection. We see the growth of a real-time detection requirement for the AML space, just as we already have in the credit fraud detection space.
Want to see what FICO is doing to help lenders get ready for compliance in 2018? Check out our financial crime pages.