This survey delves into challenges of federated machine learning beyond potential security issues that could affect adoption in industries like financial services. For example, the authors consider how asymmetric data and communications systems might make building networks between heterogenous institutions difficult and increase the costs related to uploading and downloading models or portions of models. These considerations may be especially important in underserved and emerging markets.

A recent KPMG survey of senior executives reports that the COVID-19 pandemic accelerated the rate of AI adoption across a variety of industries, including a 37% increase across various financial services uses. However, many business leaders expressed concern about this acceleration and the overall speed of adoption and welcome new guidance and regulation to foster responsible use of AI.

In the first comprehensive academic analysis of court-ordered wage garnishment, this study finds that garnishment rates nearly doubled from 2014 to 2019 due largely to increases in student loan collections. On average, 11 percent of gross earnings were remitted to creditors, raising concerns about whether unexpected garnishments may perpetuate a cycle of debt. Garnishments are most frequent in neighborhoods with higher percentages of Black residents and fewer college-educated workers, even controlling for income.

Updating a previous CFPB report, this blog documents increasing signs of financial strain through September 2022 among student loan borrowers who are expected to face monthly payments when federal forbearances end. It also tracks changes in other risk factors for which consumers are most likely to experience financial difficulties when the relief ends, depending on implementation of new debt cancellation proposals.

Comparing areas with and without flood insurance after Hurricane Harvey, this study finds evidence that mandatory flood insurance programs mitigated financial shocks where applicable. However, despite lender forbearances, federal disaster grants, and small business disaster loans, credit-constrained homeowners who lived in areas not subject to flood insurance mandates experienced a 20% increase in bankruptcies and a 13% increase in the share of debt in severe delinquency in flooded blocks relative to non-flooded areas.

This study updates mortgage market developments in the use of cash-flow information from bank accounts and utility, telecommunications, and rental payments history. The report highlights issues concerning data collection, standardization, and consumer protection regulation when using non-traditional financial data sources, as well as the impact of pricing, servicing, and regulation in determining whether the use of such data sources enhances racial equity.

This study analyzes five-year credit outcomes for consumers who participate in credit counseling and in some cases enroll in debt management plans (DMPs) using data from a national initiative by the National Foundation for Credit Counseling.

This paper examines the role of racial bias in contributing to disparities in consumer bankruptcy outcomes. Using a deep learning model trained on voter registration data to impute race in their analysis of bankruptcy filings between 1990 and 2022, the authors find that Black filers are more likely to have their cases dismissed without any debt relief under both Chapter 13 and Chapter 7 than White filers.

This study analyzes the aggregate impacts of mortgage forbearances and the Federal Reserve’s large-scale asset purchase program through Q3 2021 with a particular eye toward effects on different racial and ethnic groups.

The authors explore the implications of model multiplicity – the phenomenon in the development of machine learning models that produces several model specifications for a given task that differ in various ways but deliver equal accuracy.