Projects
AI-Enabled Personal Financial Agents
FinRegLab has launched a project to assess and shape the next generation of AI-enabled tools designed to help consumers manage and build their financial lives.
This report lays the foundation for a FinRegLab initiative that will explore opportunities to advance financial health measurement for small businesses. It proposes a framework and set of metrics for defining and measuring small business financial health using digital bank account data. The report also outlines how financial health measurement together with cash flow data could help CDFIs better measure their impact and create value for a broader range of financial institutions beyond CDFIs.
FinRegLab is evaluating the use of digital banking data to measure small business financial health, evaluate the effects of loan capital on business outcomes, and examine the contribution of entrepreneurship to wealth building in underserved communities.
This paper reflects discussions with several banks about key considerations and approaches when managing machine learning underwriting models for consumer credit underwriting.
Just a year after the inaugural FinRegLab AI Symposium highlighted the inflection point facing the sector and the broader U.S. economy, participants in the 2025 Symposium described concrete impacts, evolving risks, and expanding opportunities. This report highlights key themes and insights from the event.
Projects
Agentic AI in Financial Services
FinRegLab is tracking the spread of agentic AI in financial services and e-commerce. These dynamic AI systems can be structured to respond to new information and make and execute decisions without ongoing human engagement.
This market scan report traces the adoption of agentic AI systems in financial services and e-commerce, identifying critical questions about how to tap its benefits while mitigating risks to consumers, financial services providers, and the broader economy.
This empirical white paper assesses the impacts on model predictiveness and credit access of both adopting machine learning techniques and incorporating electronic bank account information (often called cash flow data) in consumer underwriting models.
This report details the experiences of mission-based lenders such as community development financial institutions (CDFIs) and minority depository institutions (MDIs) as they adopt electronic bank account data and lending platforms to increase lending to minority-owned businesses and other underserved entrepreneurs. The paper provides both a practical snapshot of how lenders manage data and technology adoption today and a broader analysis of issues that will shape their ability to serve substantially larger numbers of entrepreneurs going forward.
This empirical paper analyzes the impacts on predictive accuracy and credit access of incorporating electronic cash-flow data into small business underwriting models using data from two fintech lenders that lend to a broad spectrum of customers.