Is AI the End-All Solution to Mortgage Lending Inequities?

A new article jointly authored by the Federal Home Loan Bank of San Francisco and the Urban Institute takes a look at new research on the effects Artificial Intelligence (AI) in helping level the playing field for minorities and borrowers of color with the ultimate goal of advancing racial equity in homeownership. 

According to the authors, the Black/White homeownership gap is wider today than it was at the height of segregation in the 1960s, coming in at nearly 30 percentage points higher, but that number may be at a crossroads due to the rapid adoption of AI. 

Sourcing information from a new research paper entitled “Harnessing AI for Equity in Mortgage Finance,” explores how AI could be the solution to shifting the mortgage and homeowning industry toward promoting greater equity in the mortgage financing process and close the homeownership gap which has been perpetuated by long-standing bias (or unconscious biases) that persist in the wealth-building capabilities of people of color. 

The authors also cite a report from the FHL Bank of San Francisco called “Racial Equity Accelerator for Homeownership” which considers the benefits of AI on the mortgage finance industry—but cautions that adoption is still in its infancy, and the time to set goals, standards, and biases of this new technology is now before it’s adopted on a larger scale and amplified to every facet of lending. 

“AI and the potential disruption in its wake have been the subject of many conversations, and this latest paper underlines the importance of getting ahead of the technology and harnessing it to advance equitable outcomes in our society,” said Janneke Ratcliffe, VP of the Housing Finance Policy Center at the Urban Institute. “We cannot afford to wait and see how stakeholders will implement AI in their processes; we should act now to establish guidelines and guardrails to ensure that Black and Latino households are supported in their pursuit of homeownership.” 

Teresa Bryce Bazemore, President and CEO of FHLBank San Francisco, agrees: “AI has the potential to revolutionize processes in the mortgage finance industry, not only introducing greater efficiency and customization but also promoting greater equity in homeownership. However, it will only deliver the desired results if, and only if, data and algorithms are free of bias. If not, Black and Latino individuals and families will continue to be at a disadvantage when it comes to building equity and stability and, ultimately, intergenerational wealth. That outcome is simply unacceptable.” 

The authors of the paper gathered their information from nearly 50 individuals in the Federal Government, financial technology firms, mortgage lending companies, consumer advocacy groups, and research organizations. In the report, the authors describe AI’s penetration across mortgage application process and around the mortgage industry ecosystem. The adoption of AI was found to vary within the mortgage finance industry, with larger mortgage lenders, fintech firms, and government-sponsored enterprises already implementing AI in functions such as underwriting, property valuations, fraud detection, and marketing. In contrast, adoption is lower among smaller and mission-oriented lenders, such as minority depository institutions (MDIs) and community development financial institutions (CDFIs). 

The authors conclude, that based on their findings in the mortgage finance ecosystem, they provide three distinct recommendations: First, increased regulatory guidance is recommended to establish clear guidelines on applications and protect the rights of consumers. Second, intentional design must be the backbone of any AI model to ensure algorithms are free of bias and centered on equity. Finally, pilot programs should be implemented to test models and ensure that industry and consumer outcomes are consistent with intentions and support equity in homeownership. 

“Harnessing AI for Equity in Mortgage Finance” is the last in a series of four reports developed through a two-year, $1.5 million collaboration between the Urban Institute and FHLBank San Francisco. Previous reports examined incorporating alternative data into mortgage underwriting, mitigating the impact of student loan debt on Black homeownership, and using mortgage reserve accounts to help sustain homeownership. 

Click here to the jointly authored article in its entirety.