In this podcast episode, Len Suzio and Dean Stockford discuss the unregulated use of Artificial Intelligence (AI) in the banking industry. Dean brings up how, amidst the focus on fair lending and Unfair, Deceptive, or Abusive Acts or Practices (UDAAP), there have been increasing concerns about potential bias in AI systems. Specifically, red flags have been raised about scoring systems built into the lending process and the potential for inadvertent redlining in marketing systems. Both Len and Dean concur on the importance of understanding and regulating the use of AI in banking systems.
Dean suggests several proactive measures for financial institutions (FIs) in anticipation of regulatory exams. He recommends conducting an enterprise-wide inventory of all AI-utilizing systems and understanding their specific applications, starting with BSA/AML and Fraud departments, followed by Lending/Underwriting, Marketing, and Human Resources. In addition, he urges FIs to conduct fair lending data analyses and risk assessments with specific emphasis on AI in lending/underwriting systems. To conclude, Dean emphasizes the need for management to understand underwriting systems through thorough testing and analysis, and document vendors' explanations and testing results, to gain a comprehensive view of the risk these systems pose to their institutions.