Over the past several years, artificial intelligence (AI) has captured the imagination of the financial services and insurance industries, and for good reason. The enterprise has been deluged by a monsoon of data, and this data is the fuel that will feed the higher-order insights and knowledge that AI will reveal. Given the previously unimaginable compute power and storage that is now available through public cloud and the ever-shrinking cost of those resources, the future for AI in financial services and insurance is seemingly limitless.
However, a different reality confronts users when they examine current processes and systems through the lens of AI. Take, for example, the need for improved Know Your Customer (KYC) and Anti-Money Laundering (AML) tools. Simply put, as powerful as AI may be, it is generally not well suited to being slotted in as a direct replacement for existing KYC/AML solutions. Legacy KYC/AML organizations usually operate by relying on the generation of rules-based alerts rather than the creation of predictive insights. Advancing the organization to not only monitor and detect trailing indicators of fraud but expanding predictive focus to new indicators or anomalies that are “worth” monitoring is a significant undertaking.
For instance, organizations must advance their data provisioning and management capabilities, new technologies, and data science delivery techniques need to be implemented, and newly deployed capabilities must be integrated with legacy processes and systems. Additionally, the business case for the transformation of risk and compliance has historically been proven to be difficult for organizations to develop and evaluate. In the end, it is all too easy for well-intentioned KYC/AML AI efforts to fail to live up to expectations due to these factors.
In light of this, a better approach is to introduce AI to existing workloads along with a well-designed game plan for driving continuous improvement and sustained gains. Rather than going for a home run with a lift-and-shift approach, long-term, meaningful gains are best achieved through a staged approach that focuses on short-term gains coupled with long-term vision. Once momentum is established, an organization is ready and better able to roll out a larger and transformative program.
KYC and AML Really is a Big Deal
It would be difficult to overstate the magnitude of the importance of KYC/AML for financial services firms, both in terms of the current impact and the scope of future challenges. Recent analysis published in Compliance Week found that fines and penalties for regulatory violations increased by 141% between 2019 and 2020 with a total price tag of $10.4 billion last year. As a result, a recent report from research firm Facts and Factors estimated that the broader regtech solutions industry will grow over six-fold between 2019 and 2026, from $5.3 billion to over $33 billion.
The double whammy of rising fines along with dramatic increases in the cost of KYC/AML solutions might seem like enough impetus to consider AI as an alternative to existing solutions but experience shows that action towards a solution often isn’t begun until an expensive problem occurs. However, understanding the broader context and the myriad ways that process and practice can be improved and how these lead to desirable outcomes go a long way to revealing how embracing AI solutions at this point in time makes sense. Not the least, it is likely that a robust AI competency will soon become table stakes in financial services and firms that lag behind may find themselves in highly undesirable competitive environments.
Application Principles for AI in Financial Services and Insurance KYC/AML
In large part, existing KYC/AML solutions are rules-based and, therefore, both difficult to scale and inflexible when it comes to interfacing with other systems or alternative forms of data. Like so many systems, they were built this way due to the fact that they were single-point solutions to start and are the product of old-line development principles and practices. In general, they are brittle, making augmentation, rather than wholesale replacement, the best way to enhance them.
A move toward AI in KYC/AML allows for the introduction of principles-based solutions which are simultaneously more flexible, easily integrated with other systems, open to evolution and growth, and, ultimately, less expensive. Simply put, adopting AI allows for a shift from mechanically chasing alerts to generating insights through intelligent anomaly detection. This is particularly useful because the nature and substance of bad actors is continually evolving and becoming more sophisticated. Not only are old, rules-based systems incapable of detecting these types of fraud, they will quickly become even more out of date quickly in the face of ever more sophisticated fraud practices.
One key way to increase the effectiveness of AI for KYC/AML is to begin by focusing on smaller, well-defined projects. This approach has multiple benefits, including garnering immediate returns, building a base of experience, and complementing existing systems and processes. As a result, enthusiasm and acceptance is built without causing excessive disruption to operations that are sensitive to the overall health of the enterprise.
Importantly, the introduction and adoption of these new tools allow expert staff to concentrate on higher-value opportunities as they are liberated from simple, repetitive tasks. This not only leads to better results in both terms of regulatory compliance and financial performance but also influences team member effectiveness and satisfaction. This should lead to a virtuous cycle of improvement as new technologies spur greater productivity that deliver steady and consistent increases in performance metrics.
Ultimately, the enhanced data construction and storage coupled with the technology and process improvements described above can be propagated in ways that benefit the larger enterprise and positive return on investment from these efforts can alter the traditional formula for risk management that typically responds to the negative (e.g. losses and fines) rather than strives for gains that arise from lower costs and higher returns. Such an orientation isn’t “pie in the sky”: real-world experience shows that the broad adoption of modern technology principles generally leads to broad and deep positive results.
The AI Future for KYC/AML Is Here, But Gains Won’t Come All at Once
AI has been around for over five decades but for much of that time it was the province of academia, not business. The proliferation of data coupled with a dramatic increase in cloud compute and data storage power alongside falling costs for these services have created the perfect conditions for AI to exit academia and take root in the enterprise. Soon, broad-based utilization of AI for KYC/AML and other tasks will be table stakes in business. For now, getting to that level or performance will take some time and effort.
Maven Wave helps drive the future of financial services with innovative business outcomes, fueled by cloud, with risk top of mind. To help organizations maximize economic outcomes and advancements, Maven Wave brings a rich blend of industry-specific technological expertise, agile-integrated design, and best practices for transformation. Contact us to learn more.
Get the latest industry news and insights delivered straight to your inbox.