Tenants (and algorithms) in Common: FTC & CFPB Launch RFI

Data minimization, expanded disclosure, algorithmic transparency, and algorithm audits - may be coming to a rental application near you!

On the heels of the European Court's AG decision in SCHUFA, that stated that credit or similar scoring can, by itself, be automated decision making subject to Art 22 of GDPR, the the #CPRA call for comments on regulations on automated decision making - Federal Trade Commission and Consumer Financial Protection Bureau issue a new RFI on background screening for rental housing applicants. The questions seem to tease out topic that we are familiar with from the EU school of "automated decision making" and "legal and similarly significant effects" and specifically address the role of algorithms, data minimization, transparency and human decision making/intervention.

Questions include:

Data minimization and accuracy:

• How do landlords and property managers determine whether and to what extent to consider past criminal and eviction records in making a decision about a tenant or prospective tenant?
• Are there issues with the overall accuracy or completeness of criminal records that impact their usefulness in assessing individuals for housing or the benefits of considering them in making housing decisions?
• Are landlords or property managers requesting that prospective tenants disclose their source of income (including an intention to use housing vouchers) in applications? Why, and how prevalent is this practice?

Transparency:

• How and to what extent are landlords and property managers informing tenants and prospective tenants about their tenant screening criteria?
• Should landlords and property managers be required to disclose in advance their criteria for approving a prospective tenant (e.g., any disqualifying factors, such as poor credit history, prior evictions or criminal history), the information they will consider and the sources of that information (e.g., background information from tenant screening reports), the involvement of any third parties in the evaluation of the prospective tenant’s application (e.g., whether the landlord or property manager uses a tenant screening company to provide screening reports or to make recommendations, and the identity of that company), and any procedures landlords or property managers provide for disputing screening information or appealing their decisions?

Bias:

Do tenant screening practices have unique impacts on certain groups or communities? For example, are there unique impacts on historically underserved populations, such as Black, Indigenous, and people of color; the LGBTQI+ community (especially trans and gender nonconforming individuals); military service members; immigrants; public housing voucher recipients; renters with disabilities; or others?

Algorithmic Transparency and reliability:

• How are algorithms, automated decision-making, artificial intelligence, or similar technology (collectively referred to below as “algorithms”) being used in the tenant screening process?
• For algorithms that are being used for any of these purposes, how are the algorithms designed, developed, or otherwise created?
• To what extent is the performance of algorithms in tenant screening tested, both pre- and post-deployment?
• How are consumer reporting agencies deciding what criteria should be included in their tenant screening recommendation or scoring products?
• What steps, such as auditing, are taken to ensure that algorithms that evaluate, grade, or make recommendations about prospective tenants are not discriminating against prospective tenants on the basis of race, sex (which includes sexual orientation and gender identity), disability, or other protected class, and how are such steps carried out?
• What steps are consumer reporting agencies taking to ensure that the tenant screening recommendations provided by algorithms are explainable to landlords, property managers, and prospective tenants?
• How are landlords and property managers using recommendations or scores from consumer reporting agencies in deciding whether to rent to prospective tenants?

Human intervention:

To what extent do consumer reporting agencies allow tenants the opportunity to dispute, seek review of, or seek a non-automated alternative to the use of a recommendation, prediction, or score produced by an algorithm? To what extent do landlords or property managers re-assess housing applications following a tenant’s dispute or correction of scoring criteria or underlying data?

Comments may be filed by May 30, 2023.