AI Use in Credit Reporting and Housing

“The use of artificial intelligence in housing and consumer credit, banking and other critical financial transactions may lead to unlawful discrimination in violation of consumer protection, fair lending and civil rights laws. An Action Plan must guard against the systemic risks posed to consumers by the deployment of AI systems with the capacity to make decisions regarding these important life activities and the costs associated with the build-out of the infrastructure.”

The Problem

Decisions drawing on credit data reproduce and spread existing racial inequality, making it harder to achieve true economic equity.

Credit reports and scores directly impact Americans’ economic security and opportunity. Credit history can affect the way Americans are treated by lenders, landlords, utility companies, hospitals and employers. Having a poor credit history or a “thin file” with insufficient credit information to generate a credit score can mean a consumer will end up paying more for loans and insurance (or have trouble even getting them in the first place). Misuses of credit history are prevalent and harmful: Job seekers can be denied work based on their credit history, and the Trump administration has even proposed using credit history to determine whether immigrants should be eligible for permanent residency. Most harmfully, our credit system is built on—and continues to reinforce and expand—deep racial inequities. Generations of discrimination in employment, lending, education and housing have produced significant racial disparities in credit history. Past discrimination is baked into current determinations of creditworthiness: Credit scores and other lending algorithms disproportionately represent black and Latino loan applicants as “riskier” customers. As a result, decisions drawing on credit data reproduce and spread existing racial inequality, making it harder to achieve true economic equity.”

See the full article on the Demos.org website as well as the NCLC site.