Mounting approval for the use of AI and machine learning in the fight against financial crime

Blog

Mounting approval for the use of AI and machine learning in the fight against financial crime



The Wolfsberg Group is a global authority on standards for know your customer (KYC), anti-money laundering (AML), and counter-terrorist financing (CTF) in banking. The group recently issued five key principles that financial institutions should use to guide their approach to use of AI and machine learning in their anti-financial crime programs.

Possibly because of the landmark ruling in Holland earlier this year, which upheld the legitimacy of neo bank, bunq’s, use of AI technology in its AML processes, the Wolfsberg Group has issued five new guiding principles for financial institutions to consider when adopting and using this type of technology in their fight against financial crime.

The group talks about what can be achieved when harnessing the power of AI and machine learning, including “efficiency and effectiveness” in compliance processes. But there is also a note of caution that financial institutions should exercise diligence in how it’s used, and care in the technology that is ultimately chosen, because it is the institution that will bear the responsibility for any outcomes – be they positive or negative.




Data processing

Clearly, it is the amount of data that can be gathered and processed through automation that is both the boon and bane of the financial institution when using AI and machine learning. Algorithms can collect millions of pieces of data – some relevant, many irrelevant to a screening or risk monitoring process.

With such vast amounts of data to be processed, great care needs to be taken to choose the right technology that will filter out what is irrelevant, return what is material and important, and that won’t result in creating further exposure to risk, because there is simply too much data to interpret.

The AI platform chosen for effective and efficient support of an automated approach to anti-financial crime management must be developed around a firm’s risk appetite; it must learn from the decisions and outcomes the financial institution has made; and, crucially, it must not inadvertently become responsible for unconscious bias, resulting in poor decisions and people being excluded from the economy.




Wolfsberg's five principles for use of AI and Machine Learning

  • Legitimate purpose - You should ensure the technology is only collecting the data needed for an anti-financial crime process i.e., it should only collect data and use it for risk monitoring and screening processes to control CTF, AML, and KYC compliance.
  • Proportionate use - The group suggests that this form of risk monitoring does not need to be applied to everyone and for all KYC and AML activity. Instead, you should apply it to cases where the risk-threat or severity of financial crime is at its highest.
  • Design and technical expertise - Choose your providers wisely, because you need to be sure you understand what the AI can achieve, what it can’t, and the outcomes of that. This is particularly important when it comes to ensuring a system doesn’t become riddled with unconscious bias, which can result in people being excluded from the financial system unfairly.‍

Machine learning and AI are helpful for doing the bulk of data collection work, and it can support decision-making, but understanding the nuances of humans, their behavior, and their risk levels can be missed by machines. Bringing people into the process at the right time to review the data and make decisions may be optimal.

There can also be issues created through the number of “false positives” an algorithm can create.  “Solutions, like Moody’s Grid and AI Review, flag risk alerts to people for enhanced due diligence, having reduced and removed many of the false positives generated by gathering millions of pieces of information during a screening process.” said Hugo Veazey, Industry Practice Lead at Moody’s Analytics KYC.‍

  • Accountability and oversight - Whether using an in-house or outsourced solution, you will be accountable for the operation and outcomes of any AI technology used in anti-financial crime processes. That means if the AI tech goes awry, it’s on you, and this means the design, setup, and management of the system chosen is extremely important.

Choosing the right provider with a history of risk management, an understanding of compliance globally, and a suite of innovative technology solutions is the right answer. The solution should also be flexible, so it can be adapted when and if further efficiencies can be found in an AML or CTF process.

  • Openness and transparency - Finally, the group advises that you be open about how and why you are using AI; to what purpose; and about the role it plays in management of your KYC, AML, and CTF processes. But it also says, don’t give too much away, so that criminals can use the information to manipulate your systems and controls.



Moody’s Grid and AI Review

Protect your business from association with financial crime by harnessing the power of Moody’s Analytics AI Review. This product uses the capacity of artificial intelligence to complete KYC verification and AML processes.

You can integrate Review’s comprehensive, flexible, and precise third-party screening into your compliance processes. Monitor results related to individuals and entities according to your risk profiles and appetite, and make decisions with confidence about whom you do business with, and who you don’t.




About Moody’s Analytics KYC

Moody’s Analytics KYC is transforming risk and compliance; creating a world where risk is understood, so decisions can be made with confidence.

Our customers build their own unique KYC eco-system from a flexible menu of world-class products. We offer workflow orchestration, award-winning datasets, analytical insight, and integration with global providers to create powerful, digital risk management solutions.

Harnessing our innovative technology and industry expertise, Moody’s Analytics automates accurate screening and swift onboarding of individuals, entities, and third-party suppliers. Then supports perpetual monitoring of counterparty risk across any business network in near real-time.

Moody’s Analytics KYC is helping customers automate onboarding journeys in 197 countries, across 211 jurisdictions; completing 45+ million new customer checks each day; and adding 17,000+ sanctioned entities to our databases.

Our configurable solutions empower risk and compliance professionals to deliver compliance efficiency and great customer experiences - no compromise.

Get in touch to discuss how Moody’s Analytics can support automation of your KYC, AML, and CTF processes.