FinRegLab attended ICML’s Responsible Decision Making in Dynamic Environments workshop and presented our work on a subsection of our white paper. The workshop poster and presentation focused on how model diagnostic tools affect lenders’ ability to manage fairness concerns related to identifying less discriminatory alternative models used to extend credit.

“The report “examines the central role of finance in the economic recovery from COVID-19. Based on an in-depth look at the consequences of the crisis most likely to affect low- and middle-income economies, it advocates a set of policies and measures to mitigate the interconnected economic risks stemming from the pandemic—risks that may become more acute as stimulus measures are withdrawn at both the domestic and global levels.”

FinRegLab and Professors Laura Blattner and Jann Spiess of the Stanford Graduate School of Business have released new research on “Machine Learning Explainability and Fairness: Insights from Consumer Lending.”

FinRegLab, in partnership with the U.S. Department of Commerce, the National Institute of Standards and Technology (NIST), and the Stanford Institute for Human-Centered Artificial Intelligence (HAI), are hosting a symposium, bringing together leaders from government, industry, civil society, and academia to explore potential opportunities and challenges posed by artificial intelligence and machine learning deployment across different economic sectors, with a particular focus on financial services and healthcare.

“Careful, use-case specific research to understand how AI/ML with new data may affect consumers is essential to getting the rules of the road right in terms of how we regulate to protect people while making sure the benefits of the more complex analytics are trustworthy, inclusive, and beneficial.”

“Prospective borrowers with less wealth and little credit history are now deemed riskier by the automated underwriting systems that dominate mortgage lending these days. As a result, they tend to be denied more often or given higher interest rates … despite the fact that they might well be capable of responsibly making mortgage payments.”

“But with tens of millions of consumers left out of traditional credit scoring and the pandemic exposing potential problems in the current system, [lenders] are collecting and crunching all manner of other data to determine who ought to get a loan and how much they should pay.”

“We’re seeing the need to be smart and effective in this moment to make sure that the recovery is stronger and broader than after the last financial crisis.”

“Non-traditional data is an important way to give consumers access to better rates and open up borrowing to more consumers and businesses. This data can include things like rent and cell phone payments, giving lenders a broader range of information to consider.”

Coordinated action is critical between federal regulators to continue moving the growing ecosystem for customer-directed transfers toward adoption of safer technologies and practices without undermining consumers’ § 1033 rights or frustrating the law’s potential benefits for competition and innovation.