IBM RegTech Innovations

2019: An artificial intelligence-enhanced AML odyssey

Share this post:

Welcome to the brave new world of regulatory technology.

On December 3, 2018, the Board of Governors of the Federal Reserve System, the Federal Deposit Insurance Corporation, the Financial Crimes Enforcement Network (FinCEN), the National Credit Union Administration, and the Office of the Comptroller of the Currency issued a joint statement encouraging the use of innovative technologies including artificial intelligence (AI) to combat money laundering.

Highlights of the regulator’s guidance

In the statement, regulators have provided leeway to create innovative new approaches. They agreed not to penalize banks who undertake an innovative “pilot program” even if the program proves unsuccessful. Moreover, they promise that banks “will not necessarily” be punished if an innovative program identifies a gap or flaw in the existing Anti-Money Laundering (AML) framework. If not for this statement, institutions may have unintentionally exposed themselves to regulatory scrutiny while attempting to make improvements to their programs.

The statement went on to say that institutions “should prudently evaluate whether, and at what point, innovative approaches may be considered sufficiently developed to replace or augment existing [AML] processes.” This suggests that in most circumstances, banks should implement innovative pilot programs in addition to, rather than instead of, existing AML protections and protocols. This will also increase the speed at which institutions can iterate, since these pilots can run in parallel, rather than attempting to update or replace while running an AML program.

Regulators also encouraged “early engagement” from banks to discuss contemplated pilot programs for innovative AML processes. This will afford an opportunity to understand a given bank’s contemplated approach and to provide feedback. This pseudo-approval process also will afford banks an opportunity to avoid missteps should the program come under future scrutiny.

A new approach to AML

The main conclusion that can be drawn is: the status quo isn’t working for anyone. Regulators are drowning in voluminous amounts of low-risk or irrelevant Suspicious Activity Reports (SARs) just as the firms filing them are. Therefore, regulatory approval, which was once a hindrance, is now a call to action. Now that banks have the regulatory approval to start experimenting in new technologies, already common and successfully utilized, in other industries. Three applications will have the greatest impact on the way institutions approach this new AML quest:

  1. Automate alert triage: Perhaps the greatest impact to the financial crime and AML investigation and decisioning process will be the ability to fast-track or even automatically close alerts which have an unsustainable high-false positive rate. For many institutions, these represent well over 90% of their AML workload, and don’t even capture all suspicious activities.
  2. Accelerate decision making: While many financial institutions have already invested in generic automation tools, adding a layer of intelligence on basic data gathering and analysis will accelerate investigations and provide analysts with the insights upon which to make better informed decisions.
  3. Make customer risk more transparent through collaboration: New regulatory openness will also enable a new approach to customer risk ranking. One that includes information sharing among peers, technology partners and regulators making the process more transparent and efficient. New technologies such as blockchain can help facilitate a consortium built on data distribution and trust among the participants.

Your AI journey starts with IBM

AI is not just a singular approach, let alone a few capabilities. To truly replicate the complexity of human understanding, which is the original promise of AI, requires a plethora of cognitive techniques with advanced statistical models and complex algorithms, each suited – individually or in combination – for specific use cases.

Mastering each technique takes enormous resources and time. Only the most adroit institutions can master more than a handful. Therefore, majority of firms who offer cognitive capabilities, use only a few technologies. But, as the old saying goes, if all you have is a hammer at your disposal, then every problem you face becomes a nail – no matter how unique or complex – creating inconsistent and non-optimum results. As a leader in AI (in 2017, IBM was granted 1,400 AI patents), IBM uses a plethora of technologies to solve complex and evolving problems such as investigating potential money laundering events.

Importantly, the regulators did not dictate a specific approach or technique in their letter, giving banks the latitude and ability to experiment with all manner of technologies. The “fail fast” mantra has never had a place within the financial services industry, but with this regulatory directive, we can end the frustration and inaction that has brought us to this place. It’s going to be an exciting year.

RegtalkIntroducing the IBM RegTalk webinar series

Join me, along with Alistair Rennie, General Manager of IBM Watson Financial Services, and Nikhil Aggarwal, Director of Promontory Financial Group, as we discuss the impact of this new directive as well as what IBM is doing today to help financial institutions experiment with and redefine their AML approach.

Register here for the IBM RegTalk webinar series.

More IBM RegTech Innovations stories

Why shell companies are so risky (and hard to spot)

Shell companies and the hidden threat of entity risk While not inherently illegal, shell companies have been getting a lot attention recently for the role they play in illegal activity. This makes sense, as two of the legitimate uses of shell companies are the ability to shield owners from litigation and as a vehicle for […]

Continue reading

Why QIIB trusts IBM Safer Payments for cross-channel fraud prevention

Fraud prevention is about who you can trust. For financial institutions, it’s about understanding the relative risk of a customer, a merchant and/or a transaction, as well as hundreds of different factors including location, amount, device, etc. But for customers, both actual fraud attacks as well as incorrectly blocked legitimate transactions represent a breach of […]

Continue reading