The European Money Mule Actions (EMMA) report is a coordinated effort across 26 countries and includes help from Eurojust, INTERPOL, the European Banking Federation (EBF), Microsoft, and the FinTech FinCrime Exchange. In 2021, EMMA 7 arrested 1,803 suspects and led to over 18,000 money mules being identified.
Money mule rings are part of a wider chain of money laundering that costs global economies $2.7 trillion according to the 2020 FACTI report.
Money mule rings are complicated, obfuscating an illegal financial transaction by moving illicit money through legitimate bank accounts. This leaves a detection gap in mule-assisted money laundering detection. Next-generation data analytics tools can plug this gap, here is the type of complexity these tools are up against and how they can spot a money mule in a digital haystack.
The complex web of the money mule ring
The legitimate face of money mule rings
Money mules are often those you would least expect - it is this legitimacy that is used to great effect by money launderers. An example was a 38-year-old woman, who worked in a UK supermarket. Outside of her day job she worked as a money mule for a “huge fraud gang”. Another, Laura, 21, was recruited on social media and asked to transfer £700 ($900) via her bank account to a cryptocurrency platform. The gangs behind the money mule scams specifically set out to recruit people who are less likely to ring warning bells and already have verified bank accounts. This makes detection of ongoing massive money laundering fraud, complicated.
Money laundering fraudsters use obfuscation tactics to hide in plain sight. To achieve this, fraudsters use a variety of techniques to layer the movement of money; breaking up large amounts of illicit monies into smaller amounts is one of the tactics used. By using a legitimate person to take this money, and continue the chain of movement, via a verified account or verified transaction of some kind, the obfuscation can continue undetected.
Money launderers also look for jurisdictions where anti-money laundering rules are not as stringent or enforced. This convoluted, multi-stage, cross-border, operation is cleverly thought out and makes the detection of fraud all the more complex.
Can money mules be automated?
The banking and financial sector is increasingly automated, the fraudsters are following this automation trend and the industry should be aware of the likelihood of automated fraud activity. Anywhere that is open to automation in the fraud chain is likely to be used. For example, automated credential harvesting and open banking automation. The European Payments Council “Payments Threats and Fraud Trends Report 2021”, points out the use of botnets for the automation of credential theft. The report also highlights the use of automated banking processes such as Request-to-Pay being taken advantage of by fraudsters. Automation of money mule activity may exploit automated payment methods such as Authorized Push Payment (APP) to transfer money to a fraudster’s account.
Emerging devices such as embedded and open finance enabled through protocols such as Open Banking have the potential to move illegitimate money, in plain sight. The automation of money mule activity is further obfuscating the illegal money movements that money laundering depends on. Also, automation is behind faster payments, closing the window of opportunity to detect a fraudulent event.
Next-generation data analytics tools take on complex fraud that includes money mules.
Next-generation money mule gangs need a next-generation approach to detect and prevent money laundering. But how do you find a crooked needle in a digital haystack?
Next-generation anti-fraud tools take on money mules
The complexity of modern-day money laundering, facilitated by multi-part steps and money mule gangs makes for detection challenges. Blind spots, automated money wires, and broken chains, as well as the fast adaptation tactics of the money mule fraudster, require smarter analysis of the vast amounts of transaction data associated with these complex mule-assisted chains.
Next-generation data analytics holds the key to surmounting this detection challenge.
A chain of activity can be identified across the mule network, but only if using intelligent technologies that can tease out any unusual activity. These patterns of activity often have a historical basis. Machine learning algorithms can be used to look through historical data sets and identify connected accounts and unusual behavior between accounts. A spider's web may look complex, but it has patterns within its construction. Money mule-assisted fraud also has patterns, but they are deeply hidden. Anti-fraud platforms that utilize smart analytics such as PaymentGuard, apply machine learning (ML) algorithms to locate these patterns.
Pattern analysis using ML can follow the tracks of fraudulent activity and find trends in both structured and unstructured data. An effective ML-enabled data analytics engine must be multifaceted to capture fraud activity, this includes visibility across systems using a layered approach to data analysis. Areas that a next-generation data analytics platform covers must include:
- Historical data analysis: historical patterns of behavior are explored and used to create a baseline that sets the scene for anomaly detection.
- Machine learning: additional data from existing known fraud cases can be used to train the ML algorithm.
- Anomaly detection: multi-variable data analysis, across geographic location, transactions, and device information builds up a profile.
- Threat models: data from current and emerging threat models us used to feed ML algorithms, as a basis for analysis.
- Real-time monitoring: adds a layer of analysis based on real-time transaction data.
- Watchlists: known money mule accounts feed into the data analysis.
- The result is a rich seam of relevant and actionable information, used to prevent, detect, and respond to fraud.
Riding the money mule into history
Money-mule-assisted money laundering works because fraudsters use complex webs of interaction generated by fraudsters to obfuscate transactions. Automated money transfer techniques only make the situation more complicated, as it speeds up processing times, closing windows on traditional detection: to fight fire with fire, anti-fraud data analytics models must use a multi-layered approach to threat detection.
Machine learning and smart data analytics use supervised and unsupervised models to scour existing fraudulent cases, current customer behavior, and previously unseen trends to detect payment fraud and stop it in its tracks. The complex cat and mouse game that fraudsters play with FIs requires smart tools that can tease apart the obfuscation tactics. A comprehensive approach to money mule-assisted money laundering detection and prevention will beat fraudsters at their own game.
Learn more about Eastnets' robust, real-time and multi-channel fraud prevention solution.
Contact us today for a free demo.