Brighterion CEO: 2018 Was The Year Of Artificial Intelligence

Brighterion CEO: 2018, the Year of AI

Dr. Akli Adjaoute, CEO of Brighterion, wrote this AI-focused piece as part of our 2018 year-end eBook

On Dec. 3, 2018, the U.S. Treasury’s FinCEN and Federal Banking agencies issued a joint statement encouraging innovative industry approaches to combating money laundering, terrorist financing and other illicit financial threats.

As a result, anti-money laundering (AML) has been occupying the headlines as of late. The financial industry has paid $321 billion in fines just through the end of last year, as estimated by Boston Consulting Group. JPMorgan had to pay more than $2 billion in fines due to violation of the Bank Secrecy Act, tied in part to the infamous Bernie Madoff scheme.

Regulators admit that legacy AML solutions failed and they are encouraging banks to implement newer, more innovative approaches in this area.

“As money launderers and other illicit actors constantly evolve their tactics, we want the compliance community to likewise adapt their efforts to counter these threats,” said Under Secretary of the Treasury for Terrorism and Financial Intelligence Sigal Mandelker.

Legacy AML Approaches Are Ineffective

Traditional AML systems, based primarily on rules, fail to detect suspicious activities as they suffer from several key limitations when attempting to identify money-laundering activities. These limitations lead to most institutions facing the following common issues:

  • High rate of false positives: Over 99 percent of alerts are false positives, which renders legacy systems useless.
  • Corporate silos that are narrowly focused, which increases the risk of undetected suspicious activity.
  • High IT costs.
  • Non-adaptive: Money launderers change their modes of operation frequently. If one method is discovered, activity will switch to alternative methods. Business rules are not adaptive, and thus need to be frequently and manually updated to remain current.

Next-Generation Compliance and AML Solution

The Financial Action Task Force (FATF) requires that countries, competent authorities and financial institutions identify, assess and efficiently respond to the money laundering and terrorist financing risks and have the appropriate measures to mitigate them effectively.

The application of a combination of rules, fuzzy logic and artificial intelligence (AI) technologies, particularly unsupervised learning, will help efficiently meet existing and new regulatory challenges to successfully combat money laundering, terrorist financing and other illicit financial threats.

Unsupervised Learning

As historical data related to money laundering is scarce and unreliable, it is vital to utilize unsupervised learning technologies, which have the ability to gain insight from data without any prior knowledge of what to look for. Unsupervised learning is learning from unlabeled data, where particularly informative privileged variables or labels do not exist. As a result, the greatest challenge is often to differentiate between what is relevant and what is irrelevant in any particular data set.

Unsupervised learning also encompasses dimensionality reduction, feature selection and a number of latent variable models. While first-pass solutions often use business rules, the combination of these painstakingly tested and verified rules with the power of unsupervised learning technology empowers these initial solutions with far greater accuracy.

Unsupervised learning platforms utilize temporal clustering, link analysis, associative learning and other techniques to allow customers to track transaction volatility, entity interactions, behavioral changes and more.

The power of unsupervised learning for detecting money laundering shines when data from a multitude of sources can be ingested by the system. Having a system flexible enough to accept multiple data points across a variety of sources is essential in tracing the full behavior of the individuals and the assets laundered. For example, in the context of a wire transfer, first is the transaction layer securing individual transactions such as currency deposits/withdrawals, wire transfers and checks. Second is the individual or account layer; multiple transactions are associated with specific individuals and bank accounts. Third is the business or organizational layer. Fourth is the “ring” layer, which involves multiple businesses, accounts and individuals in a money laundering scheme.

Smart Agents

Smart-agents technology is a personalization technology that creates a virtual representation of every entity and learns/builds a profile from the entity’s actions and activities. In the compliance industry, for example, a smart agent is associated with each customer, merchant, terminal, etc. The smart agents associated with an entity learn, in real time, from every activity to enrich the knowledge the system has of a single customer, learning from their specific and unique behaviors over time. There are as many smart agents as active entities in the system. For example, if there are 200 million checking accounts, there will be 200 million smart agents instantiated to analyze and learn the behaviors of each account. Multi-dimensional smart agents can also be created to monitor, for example, card activity across specific merchants. Smart-agents technology allows decision-making to be specific to each checking account and no longer relies on logic that is universally applied to all customers, regardless of their individual characteristics. The smart agents are self-learning and adaptive, since they continuously update their individual profiles from each activity and action performed by the entity.

Smart agents do not rely on pre-programmed rules and do not try to anticipate every possible scenario. Instead, smart agents create profiles specific to each entity and behave according to their goals, observations, and the knowledge that they continuously acquire through their interactions with other smart agents. Each smart agent pulls all relevant data across multiple channels, irrespective of the type, format, and source of the data, to produce robust virtual profiles. Each profile is automatically updated, in real time, and the resulting intelligence is shared across the smart agents. This one-to-one behavioral profiling provides unprecedented, omnichannel risk analysis.

Conclusion

Institutions are actively trying to avoid being front and center of headlines for money laundering and terrorist financing activities, which result in huge fines, reputational damage, attention from regulators and the loss of partners and clients. As such, institutions must utilize a combination of the benefits of existing rules-based systems augmented with unsupervised learning techniques and the unique capabilities of smart-agents technology. The result is a comprehensive solution that is intelligent, self-learning, and adaptive, and will efficiently combat money laundering, terrorist financing and other illicit financial threats.