New legislation aimed at regulating artificial intelligence (AI) could have major implications for mortgage lenders using automated tools in their credit decisions. The AI Civil Rights Act, introduced by Sen. Edward Markey (D-Mass.), would outlaw the use or distribution of AI-driven solutions that may perpetuate discrimination and bias.
The potential for unintended bias to influence AI decision engines has long been a concern for compliance and technology professionals in the real estate finance industry, as well as for lawmakers and civil rights advocates.
“Whether on the Senate floor or around the dining room table, artificial intelligence is the hottest topic of the year,” Markey said in a statement. “But these complex algorithms have a darker side as well — one that has real consequences for everyday people, especially marginalized communities. I am introducing the Artificial Intelligence Civil Rights Act to ensure that the AI Age does not replicate and supercharge the bias and discrimination already prevalent in society today.”
The bill, co-sponsored by Sen. Mazie Hirono (D-Hawaii), addresses growing concerns about the potential disparate impact of bias in AI systems used in credit evaluations, employment and other key sectors. If passed, the legislation would mean legal consequences for decisions resulting in disparate impacts based on protected characteristics, while also requiring independent audits to ensure fairness before and after deployment.
According to Markey’s statement, the AI Civil Rights Act would:
- Regulate algorithms involved in consequential decisions, such as those that impact people’s rights, civil liberties and livelihoods, including employment, banking, health care, the criminal justice system, public accommodations and government services;
- Prohibit developers and deployers from offering, licensing or using covered algorithms that discriminate based on protected characteristics or that cause a disparate impact;
- Require developers and deployers of covered algorithms to complete independently audited pre-deployment evaluations and post-deployment impact assessments to identify, evaluate and mitigate any potential biased use or discriminatory outcomes;
- Require developers and deployers to mitigate any harms identified by the pre-deployment evaluations and impact assessments and ensure that any covered algorithm performs reasonably well and is consistent with its publicly advertised purpose;
- Increase transparency around the use of covered algorithms in consequential decisions, including providing individuals a right to appeal an algorithmic decision to a human decision-maker; and
- Authorize the Federal Trade Commission (FTC), state attorneys general and private individuals to enforce the Act.
Markey stressed the bill is not an effort to stifle innovation with respect to AI.
“Make no mistake: we can have an AI revolution in this country while also protecting the civil rights and liberties of everyday Americans, we can support innovation without supercharging bias and discrimination, and we can promote competition while safeguarding people’s rights,” he added.
On the other side of the discussion about appropriate legal restrictions on AI tools is Sen. Patrick McHenry (R-N.C.).
In March, McHenry tried for a fourth time to get Congress to consider legislation to create “regulatory sandboxes,” allowing financial services providers to offer innovative products and services under an alternative compliance plan, which would apply to developing AI tools. McHenry asserted his bill would “give entrepreneurs an opportunity to test legal and regulatory waters before taking new products and services to market” and “help ensure the United States continues to lead the world in financial innovation.”
Federal regulators put companies on notice in May 2023, releasing a joint statement saying discriminatory outcomes driven by innovative technology solutions, including AI tools, would not be tolerated. The statement was signed by the Consumer Financial Protection Bureau (CFPB), FTC, the Civil Rights Division of the U.S. Department of Justice, and the Equal Employment Opportunity Commission. The agencies expressed the need for a heightened emphasis on transparency concerning how automated systems make credit determinations.
Shortly after the CFPB published guidance on using AI in credit decisions in August last year, Garris Horn LLP Senior Partner John Levonick told Dodd Frank Update the move could be viewed as the bureau “setting the stage for the successful use of emerging technology.” The Biden administration issued guidance to all federal agencies on regulating AI innovations in November.
“The subtext is that the CFPB is affirmatively providing guidance on how to successfully use AI by defining what are not acceptable practices (i.e. exclusive use by the lender, through the use of AI, of only the templated denial reasons in the model forms),” Levonick said. “The best thing for emerging technology companies is to have a clear understanding of the regulatory rails defining how the technology is required to perform for their clients for the financial institutions to adopt successfully and confidently.”
The CFPB was integral in the formation of the Global Financial Innovation Network (GFIN), established in 2018 to facilitate dialogue between regulators and innovators regarding emerging technological developments, including AI tools.
In 2019, the Federal Deposit Insurance Corp., the Office of the Comptroller of the Currency, the Securities and Exchange Commission and the Commodity Futures Trading Commission joined GFIN along with 46 other authorities all over the world.
The AI Civil Rights Act is endorsed by the Lawyers’ Committee for Civil Rights Under Law, The Leadership Conference on Civil and Human Rights, Free Press Action, UnidosUS, NAACP, American Civil Liberties Union, Electronic Privacy Information Center, Public Citizen, Access Now, Asian Americans Advancing Justice – AAJ, Brennan Center for Justice, Fight for the Future, National Disability Rights Network, Common Cause, National Center for Transgender Equality, The Trevor Project, National Council of Negro Women, Encode Justice, NETWORK Lobby for Catholic Social Justice, Accountable Tech, the National Hispanic Media Coalition, Color of Change, and Writers Guild of America, East.