By now, we all know the statistics. It’s estimated that, in just one year, laundered money totals 2% to 5% of the global GDP. That amounts to a towering $2 trillion in US dollars. Traditional transaction monitoring approaches just aren’t equipped to win the money-laundering war. In fact, existing systems return false positive alerts as high as 98%. The sad truth is that outdated, rules-based strategies can’t detect money laundered through networks of criminals who deal with smaller amounts to avoid being discovered. The operative word here is “networks.” More on that in a minute.
Forward-thinking financial institutions are rapidly discovering new money-laundering detection systems driven by artificial intelligence (AI) and machine learning to rout tech-savvy criminals. These systems are proving themselves to be up to the task, providing previously unseen views of criminal behavior and detecting anomalies that traditional systems don’t see, and doing so in record time.
Why add network analytics to the mix?
AI-led and AI-automated AML systems are undoubtedly transforming financial crime detection and investigation. But the AI-driven platforms that are most revealing are the ones that also take networks into account. Network analytics is proving to be the perfect extension of AI and machine learning systems because it goes beyond looking at customer profiles and transactions. It searches for connections among related entities to reveal previously undetected relationships. In essence, network analytics expands the reach of AI-led investigations.
Here’s how the combination works
Financial Crime detection driven by AI and machine learning generates an entity alert. This alert indicates whether the flagged entity has scored a threshold against a set of models such as risk similarity, anomaly detection, change in behavior and hotspot scores.
The investigator then navigates to the flagged entity using the platform’s UI to perform further evaluation. Using network analytics, the investigator gathers supporting data about the entity as well as other entities linked by transactions (flow of funds) or demographic relationships such as addresses, mobile phone numbers, social security numbers and more.
Using both basic filtering tools and sophisticated out-of-the-box algorithms such as Social Network Analysis (SNA), the experienced investigator explores both the flagged and linked entities to conduct a more in-depth analysis to discover data patterns, thus determining risk.
Most importantly, the benefits of the combined process are far reaching. After the risk determination is made, the investigation’s outcome is fed into the AI-driven platform to help improve the models, thus continually making the platform stronger and more intuitive. A positively enforcing feedback loop is created. Using their superior pattern recognition and data supplied by the network analytics, investigators in essence train the models, thus improving the quality of alerts generated by the system’s AI and machine learning functions.
Network analytics enhance AI-driven TMS
A well-programmed network analytics tool works with your AI-driven detection platform to develop connections among groups of customers. It assists in providing holistic customer reviews by not focusing on only single entities. It identifies similarity between customer information and transaction counterparties. It fortifies prevention and earlier detection. Network analytics significantly improves the effectiveness of AI-driven AML programs. So when preparing that all-important RFI, don’t forget to ask if the company’s TMS includes network analysis as well as AI and machine learning.
By now, we all know the statistics. It’s estimated that, in just one year, laundered money totals 2% to 5% of the global GDP. That amounts to a towering $2 trillion in US dollars. Traditional transaction monitoring approaches just aren’t equipped to win the money-laundering war. In fact, existing systems return false positive alerts as high as 98%. The sad truth is that outdated, rules-based strategies can’t detect money laundered through networks of criminals who deal with smaller amounts to avoid being discovered. The operative word here is “networks.” More on that in a minute.
Forward-thinking financial institutions are rapidly discovering new money-laundering detection systems driven by artificial intelligence (AI) and machine learning to rout tech-savvy criminals. These systems are proving themselves to be up to the task, providing previously unseen views of criminal behavior and detecting anomalies that traditional systems don’t see, and doing so in record time.
Why add network analytics to the mix?
AI-led and AI-automated AML systems are undoubtedly transforming financial crime detection and investigation. But the AI-driven platforms that are most revealing are the ones that also take networks into account. Network analytics is proving to be the perfect extension of AI and machine learning systems because it goes beyond looking at customer profiles and transactions. It searches for connections among related entities to reveal previously undetected relationships. In essence, network analytics expands the reach of AI-led investigations.
Here’s how the combination works
Financial Crime detection driven by AI and machine learning generates an entity alert. This alert indicates whether the flagged entity has scored a threshold against a set of models such as risk similarity, anomaly detection, change in behavior and hotspot scores.
The investigator then navigates to the flagged entity using the platform’s UI to perform further evaluation. Using network analytics, the investigator gathers supporting data about the entity as well as other entities linked by transactions (flow of funds) or demographic relationships such as addresses, mobile phone numbers, social security numbers and more.
Using both basic filtering tools and sophisticated out-of-the-box algorithms such as Social Network Analysis (SNA), the experienced investigator explores both the flagged and linked entities to conduct a more in-depth analysis to discover data patterns, thus determining risk.
Most importantly, the benefits of the combined process are far reaching. After the risk determination is made, the investigation’s outcome is fed into the AI-driven platform to help improve the models, thus continually making the platform stronger and more intuitive. A positively enforcing feedback loop is created. Using their superior pattern recognition and data supplied by the network analytics, investigators in essence train the models, thus improving the quality of alerts generated by the system’s AI and machine learning functions.
Network analytics enhance AI-driven TMS
A well-programmed network analytics tool works with your AI-driven detection platform to develop connections among groups of customers. It assists in providing holistic customer reviews by not focusing on only single entities. It identifies similarity between customer information and transaction counterparties. It fortifies prevention and earlier detection. Network analytics significantly improves the effectiveness of AI-driven AML programs. So when preparing that all-important RFI, don’t forget to ask if the company’s TMS includes network analysis as well as AI and machine learning.
