On April 29, the Office of Fair Housing and Equal Opportunity of HUD issued two guidance documents concerning the use of artificial intelligence (AI) in housing-related practices, specifically in
tenant screening and
housing advertising. The guidance came in response to Biden's Executive Order to address potential biases in automated systems within the housing sector (covered by InfoBytes
here).
HUD emphasizes the need for transparency and fairness when housing providers and screening companies use AI-assisted tenant screening processes. The tenant screening guidance recommended best practices for housing providers and screening companies to prevent discrimination based on race, color, national origin, religion, sex, disability, and familial status. The HUD guidance provided the following six “Guiding Principles for Non-Discriminatory Screenings”: (i) choose relevant screen criteria; (ii) use only accurate records; (iii) follow the applicable screening policy; (iv) be transparent with applicants; (v) allow applicants to challenge negative information; and (vi) design and test complex models for Fair Housing compliance.
Regarding housing advertising, HUD’s guidance warned advertisers and online platforms about the risks of using targeted advertising tools that could violate the Fair Housing Act by denying consumers information about housing opportunities or targeting individuals based on their protected characteristics. To reduce the risk of violating the Fair Housing Act, HUD recommended that advertisers should: (i) utilize advertising platforms that are taking steps to manage the risk of discriminatory delivery of housing-related ads; (ii) ensure the advertisements related to housing are accurately identified to the platform; (iii) analyze the composition of audience datasets; and (iv) monitor outcomes of advertising campaigns to identify and mitigate discriminatory outcomes. Advertising platforms are recommended to: (i) ensure that housing-related ads are run in a separate process designed to avoid discrimination; (ii) avoid providing targeting options for housing-related ads that describe or are related to FHA-protected characteristics; (iii) conduct regular end-to-end testing of advertising systems; (iv) proactively identify and adopt less discriminatory alternatives for AI models and algorithmic systems; (v) ensure that algorithms are similarly predictive; (vi) ensure that ad delivery systems are not resulting in differential charges on the basis of protected characteristic; and (vii) document, retain, or publicly release in-depth information about ad targeting functions and internal auditing.
