With the evolution of technology, avenues for financial crimes like fraud, money laundering and terrorist financing, cyber-attacks, and physical attacks, have increased significantly. Financial crimes pose a threat to financial institutions both from a monetary and brand reputation perspective, and it is becoming increasingly essential for financial systems to have systems and solutions to mitigate the risk of such crimes.
The IMF estimates that the quantum of money laundering is nearly 2-5% of the global GDP and continues to grow rapidly. Canada’s Proceeds of Crime (Money Laundering) and Terrorist Financial Act (PCMLTFA) puts in place measures for the detection, prevention, and investigation of these crimes, and the Financial Action Task Force (FATF) requires financial institutions to adhere to regulatory requirements, including following KYC rules, maintaining transaction records, and reporting suspicious transactions. Canadian banks have invested in transaction monitoring systems and send suspicious activity reports (SARs) to financial investigation units; however, the current detection systems generate high volumes of false positives. Moreover, conventional AML systems are static and rule-based, and the suspicious transactions need to be manually reviewed to conclusively determine if they are indeed illicit.
There is tremendous potential for the use of AI to enhance AML systems, especially to accelerate the review process, provide additional insights, and clearing of low prioritized backlog items. Zelusit can design and build custom state-of-the-art AI solutions which can help financial institutions better detect and prevent money laundering, reduce the occurrence of false positives, and decrease workload for human investigators.
In 2019, Canadians lost close to $104 million to fraud, with credit card fraud being one of the most common types of financial fraud. Due to the wide range of financial fraud activities, financial institutions must remain diligent in identifying suspicious behavior or behavior that deviates from patterns of regular usage.
Conventional fraud prevention systems include tools like real-time credit authorization, address verification systems, card verification value, positive and negative list, etc. Some automated methods are used to recognize common attributes to discriminate transactions, but this needs to be followed by manual review, and often requires contacting the customer to verify the transaction. The high rate of false positives can significantly impact customer confidence and their relationship with the financial institution, and the use of AI can strengthen fraud management solutions by improving the robustness of the detection model and providing recalibrated rules.
This whitepaper reviews the systems financial institutions currently use to detect and prevent money laundering and financial fraud and makes recommendations on how existing systems can be strengthened using artificial intelligence (AI). It also covers the approach Zelusit takes when helping clients strengthen or build financial crimes prevention systems.