Why AI is the Only Option for Combating Money Laundering, Terrorist Financing and Other Illicit Financial Threats

This past December, the U.S. Treasury’s Financial Crimes Enforcement Network and Federal Banking Agencies green lit an initiative encouraging banks and credit unions to implement “innovative approaches” to better combat money laundering, terrorist financing and other illicit financial threats.

By recognizing that innovation within the private sector can help financial services organizations better identify and report criminal financial activity by enhancing the effectiveness and efficiency of Bank Secrecy Act/Anti-Money Laundering (BSA/AML) compliance programs, the initiative is a testament to regulators’ newfound support of emerging technologies. 

The Cold, Hard Truth: Legacy Approaches to AML Don’t Work

For too long, financial services organizations have allocated billions of dollars every year in an attempt to ensure AML compliance. Proving consistently ineffective, legacy financial compliance systems have resulted in little more than astronomical budgets and almost every bank in the world being fined for AML problems. Reports of such lapses in security have dominated the news cycle, caused stock market values to go down, led to waning consumer trust and created endless legal messes that inevitably take years to resolve and usually involve teams of expensive consultants. 

Complicating matters further, the amount of alerts generated by legacy security and compliance systems has led to widespread alert fatigue and endless false positives. In fact, research firm Ovum found that over a third of banks receive more than 200,000 security alerts every day.  No human or team of humans — no matter how talented, experienced or devoted — can keep up with such a colossal pace. 

Why Ensuring AML Compliance Has Proved So Difficult

What the U.S. Treasury’s and Federal Banking Agencies’ initiative should incite, then, is a global effort to incorporate new technologies that can guarantee AML compliance by reducing the number of false positive alerts and identifying real alerts that need to be acted upon immediately. 

Here’s the thing with AML, though: Unlike with fraud prevention, which can tap into historical transaction data to create labels that allow legacy systems to write basic rules about what’s normal and what’s not, AML doesn’t have access to historical data. Few banks want to disclose their attack data or defense scenarios, and what’s more, money laundering isn’t a simple transaction that occurs within a finite period of time. An incredibly complex and elaborate scheme, money laundering is more like a mysterious Hitchcock movie than a one-off heist.

Unsupervised Learning is the Only Worthy Opponent of Money Laundering

Given that historical money laundering data is scarce and unreliable, financial services organizations require technology that can learn from incoming data without having any prior knowledge of what to look for. The only technology capable of such sophistication is AI, or more specifically, unsupervised learning. 

By leveraging temporal reasoning, case-based reasoning and fuzzy logic, unsupervised learning technology can empower organizations to better manage their security alerts and legacy rules, proactively identify the “weakest links,” keep up with a variety of threat actions over a long period of time and automatically deliver auditor reports that detail what occurred and how. A highly adaptive technology, unsupervised learning can also detect abnormal behaviors for the first time — a critical characteristic for a scheme as unique and destructive as money laundering. 

Repetition is a rarity with money laundering, so combining as many components of unsupervised learning technology as possible to obtain collective intelligence is key. Additionally, merging unsupervised learning-powered AML solutions with other, existing tools is crucial.

Too often, fraud and AML teams are siloed according to their focus and/or geography. By sharing intelligence across any team or technology that works with risk and compliance, however, financial services organizations can accurately detect money laundering schemes within their networks and, even more importantly, help fight global threats such as terrorist financing plots.

Proactive, AI-Powered Information-Sharing is Paramount 

As long as criminals continue to exist, there will be a massive, unrelenting market for money laundering. To realistically keep up with such an active pace of illicit financial activity, financial services organizations should heed the U.S. Treasury’s and Federal Banking Agencies’ advice and incorporate unsupervised learning AML technology without delay. Not only does the unparalleled sophistication of money laundering require an equally nuanced technology, but the ever-increasing cybersecurity skills gap demands a fully autonomous, human-independent solution.

Lastly, it behooves the financial services industry to view the U.S. Treasury’s and Federal Banking Agencies’ initiative as an impetus to foster greater information-sharing. Taking inspiration from Interpol or other government security agencies, financial services organizations have an opportunity to collaborate with each other without disclosing too much sensitive information. For instance, any time suspicious activity occurs, banks could use unsupervised learning technology to instantly share an “AML Risk Score” with any other financial entity that has access to the account in question.

This level of proactive information-sharing will be essential in the ongoing fight against money laundering, and both financial services organizations and consumers stand to benefit from the timely security implications of such collaboration. demands a fully autonomous, human-independent solution.

By Dr. Akli Adjaoute, CEO, Brighterion