Filed Date: June 21, 2022
Case Ongoing
Clearinghouse coding complete
On June 21, 2022, the United States filed this suit against Meta (formerly known as Facebook, Inc.) in the U.S. District Court for the Southern District of New York to enforce the Fair Housing Act (“FHA”). The FHA prohibits discrimination in housing, including housing-related advertising, on the basis of race, color, religion, sex, disability, familial status, and national origin (“FHA-protected characteristics”). The United States alleged that Meta’s housing advertising system discriminated against its users by excluding users from receiving certain housing ads based on users’ FHA-protected characteristics. The United States sought declaratory and injunctive relief as well as monetary damages and a civil penalty. U.S. District Judge John G. Koeltl presided over this case. This is the first case the United States has filed challenging algorithmic bias under the FHA.
The United States challenged three aspects of Meta’s ad targeting and delivery system for violating the FHA: 1) Trait-Based Targeting (that Meta encouraged advertisers to target ads by including or excluding Meta users based on FHA-protected characteristics); 2) “Lookalike” Targeting (that Meta created a machine-learning algorithm to find users who look like an advertiser’s source audience, partly based on FHA-protected characteristics, and are thus eligible to receive housing ads); and 3) Delivery Determination (that Meta used machine-learning algorithms that party rely on FHA-protected characteristics to help determine which subset of an advertiser’s targeted audience will actually receive a housing ad).
On June 21, 2022, the United States obtained a settlement agreement, which was approved by the court five days later. The settlement required Meta to stop using an advertising tool that relied on FHA-protected characteristics to identify which users would be eligible to receive housing ads. Meta was further required to create a new system for housing ads to address racial and other disparities caused by its use of personalization algorithms. After reaching the settlement, Meta created a new system to address its prior algorithmic discrimination and, on January 9, 2023, the U.S. and Meta agreed on that new system’s compliance targets. Consequently, Meta will be subject to oversight and continuing compliance reviews through June 27, 2026. The parties also selected a third-party reviewer to ensure Meta adheres to the targets during the oversight period, although the court will have final say over any disputes regarding the information that Meta provides.
Summary Authors
Saba Khan (10/28/2022)
Alex Levin (9/21/2023)
For PACER's information on parties and their attorneys, see: https://www.courtlistener.com/docket/63398625/parties/united-states-v-meta-platforms-inc/
Koeltl, John George (New York)
Blain, Jennifer Ellen (New York)
Kennedy, David J (New York)
Portlock, Karin (New York)
Koeltl, John George (New York)
See docket on RECAP: https://www.courtlistener.com/docket/63398625/united-states-v-meta-platforms-inc/
Last updated March 7, 2024, 3:06 a.m.
State / Territory: New York
Case Type(s):
Fair Housing/Lending/Insurance
Key Dates
Filing Date: June 21, 2022
Case Ongoing: Yes
Plaintiffs
Plaintiff Description:
United States of America
Plaintiff Type(s):
U.S. Dept of Justice plaintiff
Attorney Organizations:
U.S. Dept. of Justice Civil Rights Division
Public Interest Lawyer: Yes
Filed Pro Se: No
Class Action Sought: No
Class Action Outcome: Not sought
Defendants
Meta Platforms, Inc. (formerly known as Facebook, Inc.), Private Entity/Person
Case Details
Causes of Action:
Fair Housing Act/Fair Housing Amendments Act (FHAA), 42 U.S.C. §§ 3601 et seq.
Declaratory Judgment Act, 28 U.S.C. § 2201
Available Documents:
Injunctive (or Injunctive-like) Relief
Outcome
Prevailing Party: Plaintiff
Nature of Relief:
Injunction / Injunctive-like Settlement
Source of Relief:
Form of Settlement:
Court Approved Settlement or Consent Decree
Amount Defendant Pays: $115,054
Order Duration: 2022 - 2026
Content of Injunction:
Comply with advertising/recruiting requirements
Issues
General:
Discrimination-area: