A recent study shows that the use of algorithms for online mortgage lending may actually help reduce discrimination against some minority groups. The study, conducted by the National Bureau of Economic Research, calls for banks to embrace more artificial intelligence tools to help end bias in home loans and lending decisions.

According to the study, mortgage lenders reject minority applicants, on average, at a rate that is 6% higher than those with comparable economic backgrounds during in-person meetings. But when applicants complete online applications, the acceptance and rejection rates were essentially the same, the study showed.

“If we look at banks, they have argued in the past that they’re in a bind when staring down potentially conflicting government mandates to reduce balance sheet risk, while not discriminating against any one demographic group,” a Forbes.com article reports.

The disparities in homeownership rates are often cited as the main culprit behind the racial wealth gap.

Homeownership among African Americans has plunged to its lowest level, at 40%, and has been gradually declining since peaking in 2004, according to U.S. Census Bureau data. From 2009 to 2015, up to 1.3 million minority loan applicants were rejected, the National Bureau of Economic Research estimates.

Researchers point to artificial intelligence as one possible way to reverse that trend. Lenders embracing AI and online mortgage platforms say they’ve seen a fivefold increase in Hispanic and African American borrowers between the ages of 30 and 40 over the last year.

Still, “the growing use of algorithms in financial services can produce results that are positive, negative, or simply unpredictable,” Forbes.com reports. “It’s important to note that 45% of the country’s largest mortgage lenders now offer online or app-based loan origination, as FinTech looks to play a major role in reducing bias in the home lending market.”