Biden admin warns lenders there is ‘no special exemption’ for AI when denying credit
The Consumer Finance Protection Bureau said that AI does not get special treatment
The Consumer Financial Protection Bureau this week warned that creditors cannot use artificial intelligence to be exempt from giving reasons for denying consumers credit — marking the latest instance of government grappling with how AI coincides with regulation.
The CFPB announced new guidance to lenders for the way in which they use artificial intelligence and other modeling when considering credit applications. The technology, which has seen a dramatic uptick in use in recent years, is being used by creditors to develop complex algorithms to make decisions.
WARREN BLASTS CLOSED-DOOR SENATE AI MEETING, CALLS FOR RAPID REGULATION
The agency said the models are often fed large amounts of data, including some information that may have been gained via surveillance.
The regulatory agency said that creditors cannot therefore deny credit or take adverse actions by pointing to broad term like "purchasing history" — instead they must reflect specific behaviors that led to the move, even if it upsets or angers customers if it not directly linked to their finances.
When issuing a denial, they must give specific reasons so that consumers’ can make better decisions, and also be protected from discrimination.
"Technology marketed as artificial intelligence is expanding the data used for lending decisions, and also growing the list of potential reasons for why credit is denied," CFPB Director Rohit Chopra said in a statement.
HERE'S WHAT GOP SEN. MIKE ROUNDS TOLD MUSK, ZUCKERBERG, OTHER EXPERTS AT CLOSED-DOOR SENATE AI FORUM
"Creditors must be able to specifically explain their reasons for denial. There is no special exemption for artificial intelligence," he said.
The agency has also made similar moves to restrict or clarify the use of algorithms, including to landlords when renting to tenants. It has also issued a rule to block what it says is potential "digital redlining" in the mortgage market via the use of "black-box models."
The guidance shows how regulators and federal agencies are beginning to work with the reality of the controversial technology. Senators met last week with tech giants, union leaders and artificial intelligence experts on how to regulate the technology.
Sen. Elizabeth Warren, D-Mass., who was instrumental in the CFBP's creation, criticized the hearing for a lack of transparency.
"I do not understand why the press has been barred from this meeting," Warren said. "What most of the people have said is we want innovation, but we have got to protect safety."
Multiple lawmakers in both chambers have introduced bills to regulate artificial intelligence and its development amid concerns on privacy, safety and the impact it may have on the job market.