2020 Journal of Community Bank Studies

2020 COMMUNITY BANK CASE STUDY COMPETITION

processing those data more efficiently and effectively” (Gibson). The benefits of AI in BSA/ AML compliance methods are limitless. C&N also hopes to see software that would produce the least number of false-positive alerts possible and accurately assign customer risk. On average, false-positive alerts constitute up to 90% of the total alerts that BSA officers must wade through to find the transactions that lead to any suspicious behavior (Gibson). Thus, machine learning is becoming more necessary as regulations become more extensive. Findings show that institutions that effectively use compliance technology spend a lower percentage on labor costs because they are dedicating less time to manually screening activity (“LexisNexis”). Although it is a pricey investment, many firms feel it is necessary to keep up with the evolving technology. Over the next few years, it can be expected that more firms will start to branch out and use machine learning and AI. Many already see it as the future and hope that it will help to save resources in the long run. Part IV: Policy Recommendations and the Future Of BSA/AML Reform Clarifying Components FinCEN and other regulators help to clarify key components of the BSA and compliance expectations by understanding the unnecessary burden the regulations put on community banks. By listening to the collective opinion of the banks, the regulators can offer their

Over the next few years, it can be expected that more firms will start to branch out and use machine learning and AI.

after its contract with Abrigo expires. The biggest innovation that would benefit C&N is a larger Artificial Intelligence/machine learning presence to address more alerts in a shorter amount of time. A study conducted by LexisNexis reports that only 25% of financial institutions have incorporated AI/machine learning into their normal compliance processes (Orunkhanov). This number is relatively small when considering how many FinCEN reports are filed per year. In 2016, banks (not including financial institutions) filed more than 700,000 money-laundering reports, up from less than 100,000 in 2012 (“SAR Stats”). It’s clear that even though the amount of reports being filed per year is increasing, the corresponding technology is growing at a slower rate. In a conference in 2018, American economist Dr. Lael Brainard stated, “AI approaches are better than conventional approaches at accommodating very large and less-structured data sets and

54