How AI improves AML efforts across the financial services industry

Money laundering and other types of financial crime have plagued the financial industry for years. Banks and other financial institutions have consistently found themselves one step behind criminals looking to take advantage of the holes within banks’ security and monitoring systems and carry out criminal activity undetected.

In response, many of these organizations have put in place anti-money laundering (AML) solutions. However, it’s no secret that these systems still leave much to be desired. Attempting to stop money laundering without concern for accuracy can create real challenges. Often normal transactions are flagged as suspicious while genuine illegal transactions slip through the cracks.

Artificial intelligence (AI) is changing this landscape altogether. It is making effective AML a reality across the industry, despite many organizations being slow to adopt the technology. With the right platforms in place, financial institutions can finally face financial crime head on, staying, if not ahead, then at least closer to lockstep.

AML efforts are changing

Most professionals in the financial services industry acknowledge that current risk reduction efforts are falling short. Platforms are notoriously faulty, generating up to 95% false positive alerts, which have a negative effect on both company processes and customers alike. In fact, AML practices have largely been the same for the last twenty years, using rule-based systems to observe transactions and raising alerts whenever system rules or scenarios are triggered.

These transaction-focused systems fail to decipher truly suspicious behavior with accuracy, even as institutions segment their users in an attempt to reduce false positives. Now, AML systems using AI have revolutionized these processes, resulting in improved functionality and accuracy in transaction monitoring.

It’s been a challenge over the past five years or so for proponents of AI to get financial institutions and regulators on board with the new technology. Many find AI difficult to understand and implement. However, institutions with an understanding and a plan for new technology have found a path forward to mitigate risk, overhaul their systems and catch bad actors earlier in their nefarious behavior.

How financial institutions can use technology to their advantage

Artificial intelligence is one of the fundamental keys to changing AML programs for the better. Tech savvy criminals are undeterred by new technology, and since they work without regulation and in smaller, organized groups, they can easily adapt and adopt new methods to achieve their goals. This leaves banks at a disadvantage — but AI can help level the playing field.

The application of entity resolution and network generation — technologies that effectively apply AI — can make a financial institutions’ risk-based countermeasures easier to understand for regulators, providing the transparency and context crucial to adhering to compliance requirements.

While a banks’ data is complex and often siloed, the use of this AI driven technology is sophisticated enough to present a true picture of risk and flag what’s most important. The evidence available today is making it increasingly clear to financial leaders that AI can help banks achieve the level of compliance that regulators expect for earlier detection of true risk, while achieving a reduction in false positives and tangible differences in the way AML monitoring works.

Within the last ten years, banks have begun deploying entity and network-based AI programs, which allows them to have a centralized approach to analyzing both internal and external data. Through these programs, many financial institutions have recognized the importance of putting individual customers in a context-based perspective, rather than simply focusing on transactions.

For example, take a transaction between a newly married person who is purchasing their first home with the help of a parent. A large, unusual transfer of funds from one account to another would likely be flagged and investigated. This would potentially result in the funds being delayed, time being wasted, and further contact being made with the customer for clarification, which could damage the customer relationship.

However, by applying contextual intelligence technology to their AI models, such as entity resolution and network generation, the bank would be able to quickly discern that the two parties are in fact related, and that the recipient was moving the funds from their account to a mortgage broker. This could then be tied to related transactions from title companies, attorneys, etc. When these pieces are put together, they give a full contextual view of the customer’s behavior, which is clearly understood by the monitoring system as not suspicious.

The context generated from this technology has the power to prevent customer frustration, help build and maintain consumer loyalty, and save time the bank would have spent on investigating an innocent transaction. Investigators know how frustrating it is to spend their valuable time delving into false-positive alerts when actual criminals are carrying out more sophisticated illicit behavior undetected.

Suffice to say that the very nature of finance leaves institutions vulnerable to crime. As criminals continue to exploit out-of-date technology, contextual intelligence is necessary to stop the feckless approach that has been the standard for too long.

Artificial intelligence is providing a solution — one that is empowering banks to better detect fraud and financial crime and keep their institutions, and their customers both safe and compliant.

Author: Clark Frogley, Americas Head of Financial Crime Solutions, Quantexa