What Are the Career Prospects for an Anti-Money Laundering (AML) Investigator?

An anti-money laundering (AML) investigator is a person who decides if financial activity is suspicious or not. Typically, AML investigators work in a top-level department at financial institutions like banks, fintech companies, money remitters, and casinos. These individuals must review customer records, open-source information, and transaction activity to make a well-supported analytical decision. 

When activity is inexplicable, makes no business sense, or is potentially related to criminal activity, investigators file a Suspicious Transaction Report (STR), known in the United States as a Suspicious Activity Report (SAR). These reports are sent to and stored by a country’s national-level Financial Intelligence Unit, often a division of the Treasury.

Let us put this in context. Money laundering is a secondary crime that has a prerequisite: money must come from an illicit source for it to be ‘dirty’ and need ‘laundering’. In the US these are called Specified Unlawful Activities (SUA) or predicate crimes. Designated SUA crimes range from drug or human trafficking to bribery and racketeering. The idea is that any dollar, peso, rouble, euro, renminbi, or yen that comes from a predicate crime is dirty. Once a person makes a transaction such as depositing those dirty funds into a bank, sending them via money remittance, or buying a used car, the laundering process has begun. Typically, that first action is known as the ‘placement’ phase of money laundering.

After placing funds in the financial system, launderers will move the money to different financial institutions, different countries, and move it through various companies in an effort to effectively ‘break the paper trail’ and distance the funds from their illicit source. This phase of the process is called ‘layering’ and also has the effect of hiding funds from taxation or, at least, making them appear to come from a legitimate source. 

When the first two phases are effective, the people at the top of the criminal enterprise end up with money that appears to have come from legitimate sources. Then, they are free to spend the money as desired. Consider that, while you might be able to buy a used BMW for $9,000 

cash, it is much more difficult to do the same with real estate or a yacht. Additionally, criminals at this level should be considered illicit businessmen, and often reinvest the funds in business operations. Bribery, drug precursor chemicals, transportation, and street-dealer labor are all expensive, after all.

To deter and prevent this money laundering process, financial institutions are required by law to have a program for monitoring, detecting, and reporting potential criminal transactions. Therefore, the AML investigator is one of several people implementing these programs as part of a systemic workflow.

Workflows in a well-designed AML program start when a customer is onboarded to the financial institution. Account opening customer due diligence (CDD) questions, by design, capture a snapshot of the customer, occupation or industry, geographical touchpoints, and likely product usage. These data points are used to assign customers a risk rating, which then decides the parameters of AML monitoring & surveillance over that relationship. 

Most financial institution customers matriculate as “low” or “moderate” risk and are onboarded directly by front-line staff. The majority of these then go through the entire customer lifecycle without ever triggering a monitoring alert of any kind. High risk customers, usually around 10% of an institution’s clients, are typically reviewed by corporate AML staff who conduct Enhanced Due Diligence (EDD) before account opening. When properly vetted and managed, high risk customer portfolios are an important and lucrative portion of any company’s client base. EDD information is usually reviewed on an annual or biannual basis, and most customers never trigger any kind of event-based monitoring or surveillance alert review.  

A small number of customers in every risk category trigger some kind of AML alert during their relationship with the institution. These alert events come from a handful of sources, including: 

• Transaction Monitoring: Computer algorithms monitor transaction activity at the account and relationship level to determine baseline behavior. This is initially based on CDD questions asking about the amount and frequency of expected transaction activity. Monitoring can be rule-based, complex behavior based, and category based, and is increasingly being improved by AI and machine learning. Current monitoring is often inefficient and results in a large volume of alerts. 
• Team Member Referrals: Human intelligence is still the best source for beginning any investigation. Customer-facing staff notice when individuals behave strangely, information doesn’t match, and when something just ‘feels off’. 
• High-risk customer surveillance: Certain customers that pose an elevated financial crime related risk will be subject to additional, constant monitoring to ensure activity is within established parameters. 

In its current form, AML work can be partially robotic, but still requires all the complex thinking skills of a college graduate. The process of digging through thousands of lines of transactions, attempting to tie public data to customer information, and making sense of it all with limited visibility is cumbersome and time consuming. The work requires highly specialized knowledge spanning niche US and international laws, banking products, and a variety of industries. Combine this knowledge with a deep understanding of criminal and terrorist network operational techniques and you begin to get a quality AML investigator. 

Yet, the robotic aspects of the job persist. To some, the growing proliferation of Artificial Intelligence (AI) tools represent a threat to the human centric investigative approach. It is true that AI will replace some of the more mundane aspects of this industrial-scale work. 

That said, while each link of the AML workflow chain can be force-multiplied via the power of AI, each is also highly vulnerable to rote oversimplification by machine rule. As has been true since the invention of the wheel, the steam engine, and the digital algorithm, machine tools still require ‘humans in the loop’ to consistently validate and improve model outcomes. So, for the foreseeable future, highly skilled workers are needed for every phase of the AML workflow.