Fed watchdog warns AI, machine learning may perpetuate bias in lending 

The Federal Reserve’s top watchdog warned Tuesday artificial intelligence and machine learning could bolster bias in lending practices.

“While these technologies have enormous potential, they also carry risks of violating fair lending laws and perpetuating the very disparities that they have the potential to address,” the Fed’s vice chair of supervision, Michael Barr, said at the National Fair Housing Alliance (NFHA) 2023 national conference.

While new artificial intelligence tools could relatively cheaply expand credit to more people, machine learning and AI may also exacerbate bias or inaccuracies inherent in data used to train the systems or make inaccurate predictions, Barr added.

The Fed recently announced two policy initiatives to address appraisal discrimination in mortgage transactions.

On June 1, the Fed and several agencies requested public comment on a proposed rule to implement quality control standards in automated valuation models. Under the proposed rule, institutions that engage in certain credit decisions would be required to adopt policies, practices and control systems that ensure a "high level of confidence" in automated estimates and protect against manipulation of data.

A week later, the same agencies invited public comment on guidance to help financial institutions incorporate “reconsiderations of value” into their home appraisal process. Valuations may contain errors, omissions or discrimination that could affect the value of the appraisal, the agencies argued, and a reconsideration of value could help mitigate the risk of improperly valuing real estate.

“Homeownership is an important way for families to build wealth, and we should give them every opportunity to share in those benefits,” Barr said, noting he was “fully supportive” of both policy initiatives.

A quartet of federal agencies, including the Federal Trade Commission and the civil rights division at the Department of Justice, announced in April their commitment to cracking down on automated systems that cause harmful business practices.

“Fair lending is safe and sound lending,” Barr said, soliciting applause from the audience.

“When have you ever heard a vice chair of the board of governors speak against algorithmic bias?” NFHA President and CEO Lisa Rice asked. “I’m telling you, I’m excited.”