Why Artificial Intelligence Technology is the Future of Financial Crime Mitigation
Tuesday, September 26, 2017
Posted by: Brian Monroe
*ACFCS Exclusive Special Contributor Report*
By David McLaughlin, Founder and CEO of QuantaVerse
September 26, 2017
Today's financial institutions face a wide array of challenges including cybersecurity defenses, customer service innovation, and an explosion of data management issues.
More importantly, they deal with a plethora of regulatory obligations and compliance requirements regarding anti-money laundering (AML), counter-terrorist financing (CTF), Bank Secrecy Act (BSA) and Know Your Customer (KYC).
Despite an increase in spending by financial institutions to meet regulatory and compliance requirements, the money laundering problem has failed to subside and the risk of penalties remains extremely high.
In fact, according to various reports, global AML spending is estimated to be in the tens of billions of dollars annually.
Not to mention banks around the world have paid approximately $321 billion in fines since the 2007-2008 financial crisis in which regulators stepped up their enforcement efforts, according to a report by the Boston Consulting Group.
In 2016, government authorities and state and federal regulators levied a reported $697 million in penalties against financial institutions for BSA and AML shortcomings.
As well, a recurring theme in many of these monetary penalties and enforcement actions is that banks broadly don’t have the right transaction monitoring systems in place, they are not tuned properly or are producing alerts far in excess of a given compliance team’s ability to analyze them.
Part and parcel of the reason why so many banks are looking at AI as well is that it could take at least some of the decision-making pressure out of the hands of AML analysts, which can be very subjective.
That would address another major criticism of regulators in AML enforcement actions that the experience of individual financial crime compliance officers, or the cumulative acumen of their team, are not in line with the overall risks the bank is facing.
As a result, those individuals, even when the monitoring system produces accurate, quality alerts, fail to make the right decision, escalate the alert and turn the information into a report to law enforcement, the end customer of the herculean compliance efforts by the financial community.
Apart from federal examiners, banks are facing growing pressure to have more advanced AML systems at the state level, with New York recently finalizing a rule requiring institutions to craft, tune and test monitoring systems to ensure they are working as intended, and have top a compliance executive, or board member, sign off on that assertion.
Will AI Turn the Tide?
Establishing and supporting effective AML programs has been elusive for financial institutions due to their reliance on legacy technology, the ever-increasing case load for AML investigators and intransigent regulatory policies.
While transaction monitoring systems (TMS) are essential for the maintenance of an effective AML program, these systems are failing to flag half of the transactions that pose potential risks to financial institutions.
There is widespread agreement that technology advancements will turn the tide against financial crimes occurring within the banking system.
Breakthroughs in the areas of data science, computational advancements and big data practices have accelerated the pace of technological innovation. They have also enabled artificial intelligence (AI) and machine learning to enhance human cognitive performance in the execution of non-routine tasks.
In recent years, AI has proven itself as a valuable technology and is now used extensively in myriad industries, including online retail, healthcare, automotive technology and law enforcement.
Seeing promise as well, financial institutions have begun “test-driving,” AI which has produced exceptional early results. The question surrounding the use of AI within financial institutions has evolved from “if” to “how” to put AI to work.
As proven in other industries and with early adopters in the banking industry, AI offers the ability to radically improve the detection of suspicious account activity.
How AI is Implemented into the AML Framework
With the help of AI, financial institutions can analyze massive amounts of transactional and client information from a variety of sources, such as TMS, KYC databases, Lines of Business (LOB) customer information, as well as investigative databases, public Internet sources and the deep web.
To conduct such analysis, AI systems utilize agents which are highly specialized algorithms responsible for collecting and interpreting data, modeling behaviors, detecting anomalies, inferring relationships, and identifying issues.
These agents then report any irregularities to a machine learning engine by delivering both the alerts and all necessary supporting evidence.
Examining voluminous bank data, AI systems “learn” common patterns of behavior and compare transactional data against those patterns to identify behavioral anomalies. For example, while there is no economic purpose for a government fire protection agency to purchase fertilizer, that business relationship would not be flagged by a TMS.
On the other hand, an AI solution can compare the NAICS (North American Industry Classification System) codes of both entities and would determine that they were not engaged in complementary lines of business.
An AI solution would trigger an alert for further investigation into the two entities and their business transactions.
AI can identify patterns and typologies in sets of data that TMS and human investigators simply don’t have the time to find. An AI system automatically pulls and consolidates data points, scores transactions for risk and documents anomalies.
With the help of AI, AML investigators can now evolve from researchers desperately fighting against the clock to unearth relevant data, into analysts presented with automated financial crime reports and evidence that allow them to make more informed and more accurate determinations more quickly.
Transaction monitoring systems are still widely used by financial institutions, but there are many inherent limitations. AI can enhance TMS to improve AML processes in the following ways:
- Identifying anomalous transactions that are not being flagged by TMS
- Identifying all transactions, including “below the line” activity – transactions below customer transaction report (CTR) and suspicious activity report (SAR) thresholds – that are currently unmonitored
- Continuously improve AML processes by identifying innovative and new money laundering techniques, as well as updating rules-based TMS
- Enhancing the ability to assess transactions or cases that historically have limited data and lead to the de-risking of entire jurisdictions
- Reviewing the flags created by TMS and scoring each in a manner that allows cases that do not warrant an investigation to be discharged, and escalating specific cases that require further human analysis
- Improving the research process by serving up the most relevant data points required for speedier and more accurate human determinations
The Positives of Lowering False Negatives
False negatives, which are illicit transactions unflagged by TMS, represent serious risk for financial institutions.
While institutions can technically be in compliance with regulatory requirements, it is not enough to prevent all financial crimes as banks are charged with detecting potentially illicit activity and reporting that intelligence to law enforcement.
One example of how a bank can benefit from AI-infused systems, to better find suspicious transactions that were missed the first time around, occurred when one of the world’s largest banks engaged our company to increase its financial crime compliance effectiveness and improve alert outcomes.
After upgrading to AI-enabled systems, the bank, which serves millions of customers around the globe, was better able to analyze current and historical transactional data to detect false negatives that may have been missed by its TMS.
An analysis of a single month’s worth of previously unflagged transaction data detected thousands of end-clients who had originated suspicious transactions, representing tens of millions of dollars of likely illicit cash flows.
One end-client, in particular, generated suspicious transactions totaling millions of dollars that were worthy of closer examination.
The results from that bank’s engagement with AI, and the markedly improved detection capabilities, is just a small glimpse of what is possible when you combine the inherent knowledge and passion of a dedicated and driven financial crime compliance community with advanced systems that allow them to reach their full potential, even in the face of shrinking resources and growing regulatory pressure.
Those in the financial crime risk landscape have been abuzz about the possibilities of AI to streamline and improve financial crime investigations, address regulation and compliance, and effectively thwart criminals from using the banking industry for their evil agendas.
Financial institutions should give serious consideration into taking appropriate action to mitigate financial crime by deploying AI technology into their AML ecosystems.
About the author:
David McLaughlin is CEO and founder of QuantaVerse, which uses data science and artificial intelligence to help institutions better identify financial crimes. He spent six years as a naval officer, starting in 1986 as an Ensign in the U.S. Navy and attending flight school in Pensacola, FL. McLaughlin is a graduate from the highly regarded TOPGUN program, and completed a combat tour in the Persian Gulf where he was awarded the Distinguished Flying Cross and two Air Medals for bravery in combat. Prior to founding QuantaVerse, David held senior executive positions with IPR International, NES Financial and SEI.