By: Sandy Lee
Algorithmic technology is rapidly reshaping the housing market, particularly through tenant screening. Landlords increasingly rely on tenant-screening algorithms to perform background checks on prospective tenants to streamline the paperwork-intensive rental process and reduce fraud and evictions.[1] While these tools are intended to evaluate tenant risk, they are proving faulty and often creating new barriers to housing – especially for minority groups.[2]
A recent lawsuit has been filed against JWB Property Management LLC (“JWB”), one of Northeast Florida’s largest landlords, by four Jacksonville residents.[3] The plaintiffs allege that JWB used flawed algorithms from SafeRent Solutions LLC to unfairly exclude Black applicants from housing opportunities.[4] JWB is a dominant actor in the Florida housing market, with thousands of homes rented and managed across Duval county.[5] Like many landlords, JWB utilizes tenant-screening technology – specifically SafeRent Solutions – to filter through applications quickly.[6] While these screening programs are designed to eliminate applicants deemed risky due to factors like financial instability, eviction history, or criminal records, it is argued by tenant advocacy groups that they are being used to circumvent fair housing protections.[7]
In the Jacksonville lawsuit, the plaintiffs allege that SafeRent’s algorithms flagged them as undesirable tenants despite having proof that their past eviction filings were either erroneous or dropped.[8] This situation exemplifies how discriminatory practices can lead to discriminatory practices: the plaintiffs may have been categorized based on flawed or biased data, with their actual circumstances disregarded.[9] Even with this evidence, JWB refused to rent to the plaintiffs, raising serious questions about how tenant-screening technology may perpetuate housing discrimination.[10] If proven true, this case could demonstrate how AI and algorithms perpetuate housing discrimination rather than objectively mitigate risk factors.
This lawsuit also highlights the lack of regulatory oversight on the algorithms being used by tenant-screening companies.[11] The U.S. Department of Housing and Urban Development (HUD) has recognized this risk and released guidance on the role of AI in tenant screening.[12] HUD emphasizes that landlords and screening companies must ensure compliance with the Fair Housing Act and underscores the need for transparency in evaluating tenant-screening policies to prevent discrimination.[13]
The outcome of this case could have far-reaching implications for the future use of AI and algorithms in tenant screening nationwide. If a ruling is made in favor of JWB, landlords may be encouraged to continue using these problematic screening services without concern for regulatory repercussions. This could potentially set a dangerous precedent, allowing businesses to prioritize profit over fairness in the housing market and perpetuate discrimination through AI-powered decision-making. Conversely, if the plaintiffs are successful, businesses may be compelled to reassess their use of automated screening tools to ensure compliance with fair housing laws.
[1] See Patrick Sisson, For Tenants, AI-Powered Screening Can Be a New Barrier to Housing, Bloomberg (Sept. 11, 2024, 10:30 AM), https://www.bloomberg.com/news/features/2024-09-11/ai-powered-tenant-screening-tech-worries-fair-housing-advocates?srnd=phx-ai.
[2] See Mapping the Growth of Eviction Record and Tenant Screening Protections, PolicyLink https://www.policylink.org/mapping-the-growth-of-eviction-record (last visited Sept. 15, 2024) (stating that “much of the information that these companies retrieve from court records are incomplete, or even inaccurate in 22 percent of eviction records.”).
[3] Byrd v. JWB Prop. Mgmt., LLC., No. 3:23-cv-266-WWB-JBT, 2024 U.S. Dist. LEXIS 98155, at *1.
[4] Id. at *2 (alleging that “Defendants implement, utilize, and enforce a policy to determine tenant worthiness that discriminates against Black rental applicants in violation of the Fair Housing Act 42 U.S.C. § 3604.”).
[5] Charlie McGee, Judge Rules JWB Must Face Trial Over Alleged Use of Algorithms Against Black Renters, The Tributary (Jun. 10, 2024), https://jaxtrib.org/2024/06/10/judge-rules-jwb-must-face-trial-over-alleged-use-of-algorithms-against-black-renters/.
[6] See Sisson, supra note 1.
[7] See U.S. Department of Housing and Urban Development, Housing Discrimination Under the Fair Housing Act, Hous. Urb. Dev., https://www.hud.gov/program_offices/fair_housing_equal_opp/fair_housing_act_overview (last visited Sept. 20, 2024).
[8] See Byrd, 2024 U.S. Dist. LEXIS 98155, at *7 (stating that “Ms. Byrd’s rental application with JWB for an apartment rented by JWB was deterred and then denied because a search of court records by SafeRent found an eviction filing against her that was erroneously filed.”).
[9] See id. at *4 (finding that “eviction histories reported by tenant screening “Consumer Reporting Agencies” such as SafeRent, routinely include information that is inaccurate, outdated, or that belongs to someone other than the subject of the screening report.”).
[10] Id. at *2 (alleging that “SafeRent or another tenant screening company compiles and reports the mere existence of eviction filings against rental applicants to JWB and JWB Capital without regard to the correctness of the filing or the case disposition.”).
[11] See Lauren Kirchner, Access Denied: Faulty Automated Background Checks Freeze Out Renters, The Markup (May 28, 2020, 5:00 PM), https://themarkup.org/locked-out/2020/05/28/access-denied-faulty-automated-background-checks-freeze-out-renters (stating that regulators can force credit bureaus and employers to follow certain standards when it comes to records and screening, but none of those restrictions apply to tenant screening).
[12] U.S. Department of Housing and Urban Development, HUD Issues Fair Housing Act Guidance on Applications of Artificial Intelligence, www.hud.gov, https://www.hud.gov/press/press_releases_media_advisories/hud_no_24_098 (last visited Sept. 20, 2024).
[13] Id.
