The Evolution of Anti-Money Laundering

February 10, 2023

Money Laundering has been around for as long as there has been trade. It is a practice almost as old as time itself. For instance, as early as 2000 BCE in China, wealthy Chinese merchants would move their profits outside the country because commercial trading was not supported by the government. They would then reinvest their smuggled funds into other enterprises. Whilst this was not technically ‘laundering’, it is the same principle used in money laundering today.

In the United States, money laundering took off in the 1920s during the prohibition era in which there was a ban on the manufacture, transportation, and sale of alcohol. It is rumoured that Al Capone, a major mob boss in Chicago set up laundromats across the city to disguise the origin of the money earned from alcohol sales. Illicit funds were then added to the revenue generated by the laundromat and so re-introduced into the financial system, effectively ‘laundering' the illegal or ‘dirty’ money. In movies like the ‘Godfather II’ and ‘The Wolf of Wall Street’, we see money laundering re-enacted: art imitating real life.

What is Money Laundering?

Money laundering is the process of making usually large amounts of money generated by criminal activity appear to have come from a legitimate source. The illegal money is considered dirty, and the money at the end, after being laundered, is considered to be clean.

Impact of Money Laundering on Society

Money laundering is currently global, pervasive and staggering in proportions. Statistics from the United Nations show that about 2% to 5% of the world's GDP is laundered every year. That's approximately $800 billion to $2 trillion, annually. The true cost of money laundering is difficult to quantify. Less than 5% of assets lost to financial crimes like money laundering are ever retrieved.

Money laundering affects businesses, economic development, and society at large. It poses a reputational risk to organisations that allow illicit funds through their systems. If illegal funds can be easily processed through a financial institution, then it could unwittingly become part of the criminal network itself. When money launderers transfer money through illegitimate businesses, they compete with real business owners.

Economies with inadequate controls are particularly vulnerable to launderers looking for new routes to launder their funds. This leads to a damping effect on direct foreign investment and becomes counterproductive as this is the very thing that such economies require.

Money laundering enables criminal activity in society to continue – organised crime can acquire control of large sectors of the economy through investment thereby weakening the economic fabric. When money launderers invest dirty money in hard-to-detect markets, they inflate those markets the wrong way.

Stages of Money Laundering

Money laundering typically occurs in three phases: placement, layering and integration:

- Placement covertly injects the “dirty money” into the legitimate financial system.

- Layering is the process by which the source of the money is concealed through a series of transactions and bookkeeping tricks, making it difficult for investigators to see exactly where the money went.

- Integration is when money enters the legal economy. At this point, the money appears clean and the now laundered money can be used by the criminals for whatever purposes they have in mind – usually undetected, as investigators can no longer see where it originally came from.

In real life, money laundering doesn’t always follow these three phases sequentially, and some stages can be combined or repeated several times.

What is Anti-Money Laundering?

Anti-Money Laundering (AML) refers to the regulations, policies, procedures and more recently, technologies put in place to prevent criminals from concealing ill-gotten funds.

Through the years, as financial crime has evolved, so has AML legislation and subsequently the AML techniques, to implement them. The 1970 Bank Secrecy Act, Financial Action Task Force (FATF), Organisation for Economic Co-operation and Development (OECD), International Monetary Fund (IMF) etc. are just a few legislative bodies involved in the fight against money laundering.

The History and Evolution of Anti-Money Laundering Legislation and Techniques

Despite all the legislation, money laundering continues to grow worldwide, and the techniques used to evade detection are becoming more sophisticated, particularly with the shift towards a digital economy. For instance, transactions are faster, move across borders and jurisdictions more seamlessly, and the anonymity of some cryptocurrencies has been taken advantage of. Criminals embrace new technology to exploit gaps in control environments, even quicker than the regulators – it is the classic case of “cops and robbers”. 

Rules-Based Scenarios vs. Machine Learning

Traditionally, AML techniques have been based on rules and scenarios. These techniques have always been a step behind financial crime criminals. Artificial Intelligence (AI)/Machine learning (ML) is becoming the cornerstone of AML solutions such as Fraudio’s AML Solution, leading to greater automation and consequently, speed and accuracy.  ML can be applied across the entire AML value chain e.g., Risk Rating, Client Screening, Transaction Monitoring, Event Driven Reviews, Transaction Filtering & Screening, and KYC Refresh.

Traditional rules-based AML systems run on engines that only use rules and therefore tend to provide high-volume, low-value alerts. AML programs powered by ML on the other hand, often utilise both rules and models. This significantly decreases the number of false positives generated and allows for more accurate risk scoring which means less time wasted on meaningless alerts. It therefore increases operational efficiency and requires less manual intervention.

ML models are trained with algorithms based on historical data to predict future behaviour and can learn and adapt continuously, lending itself to unsupervised learning. Fraudio’s AML solution employs this in detecting patterns that are different from what is expected (anomaly detection) which is what is needed for a dynamic target like AML. 

Even though, machine learning models take time to learn, making them slower to implement, they capture both good and bad patterns better than a set of rules (or a human) ever could. The result is that criminals have a harder time deceiving Money Laundering detection systems.

Suspicious Activity Report (SAR) generation is another AML process that is evolving. When something suspicious occurs within a financial institution, an investigation is carried out which traditionally entails interrogating multiple or relational databases.  Once sufficient evidence is gathered and assessed, and is seen to justify the initial suspicion, a SAR is raised and then submitted to the relevant governing body. Understandably, this takes a lot of time as money laundering is designed to be disguised and to make investigators’ lives as difficult as possible. 

Fraudio’s AML Solution moves away from this traditional ‘semi-manual’ and tedious approach, by detecting transactions that could potentially be linked to money laundering. It does this by taking transactions as input data points. Using feature engineering techniques, it puts these data points into meaningful sequences and clusters. These sequences of transactions are based on patterns such as unusually large monetary amounts, or high frequency of transactions in a short time burst, etc. ML techniques are then applied to identify anomalous behaviour and produce alerts based on this evaluation. This gets rid of the need to raise SARs ‘manually’ and saves time and effort in the process.

Adopting new ways of fighting money laundering is the way to overcome new criminal tactics, and machine learning is the way forward. Anti-Money Laundering will continue to evolve as new crimes emerge, technology evolves, and existing measures become insufficient. To this effect, technology-friendly solutions, such as Fraudio’s AML Solution are here to deal with changing ML crimes and risks.

Measure results yourself !

How about trying our solution  and experiencing the next generation for yourself?