Publication / AI & finance
Generative AI and the Evolution of Mortgage Underwriting: Measuring and Mitigating Racial Disparities
The Journal of Real Estate Finance and Economics (2026) Journal article
Auditing racial disparities in LLM mortgage underwriting, and testing ways to reduce them.
Abstract
Mortgage underwriting is entering a new phase of technological change as general-purpose generative AI tools begin to appear in mortgage origination workflows. We develop an audit framework to evaluate how large language models process mortgage applications. Using real HMDA applications, we experimentally manipulate borrower race, race proxies, and credit quality, and prompt leading commercial LLMs to make approval and pricing recommendations. LLMs recommend more denials and higher interest rates for Black applicants than otherwise identical white applicants, with disparities largest for riskier loans and persisting when race is signaled through names or locations. A simple instruction to avoid bias substantially reduces disparities when race is explicit, partially reduces the name-based pricing gap, and does little against the geographic pricing gap. In matched HMDA–Freddie Mac data, race-blind LLM recommendations correlate with actual lender pricing and subsequent delinquency, though the delinquency association operates largely through observed credit scores. The results show how AI auditing can inform mortgage underwriting governance before general-purpose models become more deeply embedded in origination workflows.
Figures

LLM recommendations are worse for Black applicants, but mitigation instructions narrow the gap

Racial disparities persist across 16 leading LLMs

LLM interest-rate recommendations track lenders’ actual mortgage pricing

Changing race in otherwise identical loan applications isolates racial disparities

ChatGPT claims to evaluate loan applications without discriminating
Links
Media & discussion
- No Broker, No Bank, No Problem: How AI Could Turn Mortgages Into a Self-Service Experience | Mortgages | U.S. News
- Telling AI not to be biased surprisingly works: a study | National Mortgage News
- National Mortgage News (PDF copy)
- Study Shows How Lenders Can Remove AI Bias – The Mortgage Note
- The Mortgage Note (PDF copy)
- As AI takes the helm of decision making, signs of perpetuating historic biases emerge
- News from the States (PDF copy)
- ilLUminate Podcast | Lehigh Business
Awards & recognition
- Best Paper Award at the 2024 New Zealand Finance Meeting
- ARES 2025 Manuscript Prize for Real Estate Finance ($1,000)
- ARES 2025 Innovative Thinking “Thinking Out of the Box” Prize ($1,000)
Notes
Formerly titled “Measuring and Mitigating Racial Disparities in LLMs: Evidence from a Mortgage Underwriting Experiment”