Artificial Intelligence and Machine Learning: The Opportunities and Challenges of Using Big Data

Abstract

In Chapter 4, “Artificial Intelligence and Machine Learning: The Opportunities and Challenges of Using Big Data,” Matthew Adam Bruckner, Associate Professor of Law, Howard University School of Law, explores the potential offered by Big Data, artificial intelligence, and machine learning in financial services. Using data from fintech lenders such as Lending Club and Upstart, several recent studies suggest that fintech lenders’ use of nontraditional (“alternative”) data present significant opportunities to greatly improve financial inclusion. For example, the Consumer Financial Protection Bureau’s analysis of Upstart’s lending data found that “the tested model approves 27% more applicants than the traditional model, and yields 16% lower average APRs for approved loans.” But a recent report from the Student Borrower Protection Center warned that fintech lenders’ use of educational data points may penalize Black and Hispanic borrowers for attending a community college, a historically Black college or university, or a Hispanic-serving institution. Bruckner explains the current regulatory environment, including how certain fair lending rules can foster innovation in credit underwriting and where the use of nontraditional data can violate these legal and regulatory protections. He also outlines key trade-offs regulators need to consider and the limits of building fair artificial intelligence/machine learning–based credit algorithms and underwriting models.