•
The Justice Department and eight states sued RealPage in 2024 over software that recommended rents using competitors' nonpublic data. The November 2025 settlement is now the closest thing antitrust law has to a rulebook for pricing algorithms.
The Justice Department sued RealPage on August 23, 2024, alleging that its revenue management software let competing landlords price apartments off each other's nonpublic data. Fifteen months later the case settled. There is still no court ruling on whether an algorithm can carry a price-fixing agreement, but the settlement terms have become the working rulebook for anyone who prices with software.
Correction, September 2026. An earlier version of this post described the litigation as a Texas-led state action captioned State of Texas v. RealPage. That was wrong. The case at docket 1:24-cv-00710 is United States v. RealPage, Inc., filed in the Middle District of North Carolina and joined by eight state attorneys general. Texas is not among them. The docket number was right. The caption and the court were not.
| Item | Detail |
|---|---|
| Caption | United States v. RealPage, Inc. |
| Court | U.S. District Court, Middle District of North Carolina |
| Docket | 1:24-cv-00710 |
| Filed | August 23, 2024 |
| Plaintiffs | United States, plus North Carolina, California, Colorado, Connecticut, Minnesota, Oregon, Tennessee, and Washington |
| Claims | Sherman Act section 1 (agreement in restraint of trade) and section 2 (monopolization) |
| Status | Proposed settlement announced November 24, 2025, subject to court approval |
RealPage sells revenue management software to apartment landlords. The product ingests data about a market and recommends a rent for each unit.
The government's allegation is about what went into the recommendation. According to the DOJ complaint as reported by ProPublica, RealPage collected nonpublic, competitively sensitive information from participating landlords, including actual executed lease prices, current occupancy, and lease expiration dates, and fed it back into the pricing recommendations delivered to their competitors. The complaint also alleged that RealPage discouraged landlords from deviating from those recommendations, and that its data advantage entrenched a monopoly in the market for commercial revenue management software.
RealPage's answer has been consistent: landlords set their own prices, the software makes suggestions, and using data to price a product is ordinary business.
In January 2025 the department expanded the case, suing six large landlords for participating in the scheme.
Section 1 of the Sherman Act requires an agreement. Parallel behavior alone is not enough. Two landlords who independently read the same market and raise rents the same week have not conspired.
The government's theory is that the software supplied the missing element. If competitors feed nonpublic pricing data into a shared tool, and the tool returns recommendations that account for what rivals are actually charging, the participants have exchanged information they could not lawfully exchange in a hotel conference room. The intermediary does not sanitize it. This is sometimes called a hub-and-spoke arrangement, with the vendor as hub.
The defense is that a recommendation is not an agreement, that landlords rejected recommendations routinely, and that better information makes markets work rather than breaking them.
The case settled before a court answered that question at trial. That is the single most important thing to understand about it. There is no merits ruling holding that algorithmic pricing violates section 1.
The proposed settlement, announced November 24, 2025, is unusually specific about product behavior. Law firm analyses from Wilson Sonsini, Fenwick, and Paul, Weiss walk through the terms. The provisions that matter:
RealPage may not use nonpublic competitively sensitive information, current or historical, in the runtime operation of any revenue management product. That is the core prohibition.
Model training is treated differently. RealPage may train models on nonpublic competitor data if that data is at least 12 months old. The theory is that year-old lease data cannot coordinate tomorrow's price.
Product features that nudge in one direction are constrained. The "Auto Accept" feature that implements recommended price changes automatically must have its range set manually by the user rather than defaulting on. The "Governor" feature, which the government alleged favored increases over decreases, has to operate symmetrically.
Whatever you think of the underlying theory, that is a workable compliance checklist, and it is being read as one across every industry that prices with software.
The settlement requires court approval, and the landlord defendants sued in January 2025 are on their own track. Private class actions brought by renters continue separately from the government case and are not bound by the settlement's terms.
The larger question is unanswered and will stay unanswered until a court reaches the merits in some algorithmic pricing case. Until then the settlement is the de facto standard, which is a strange way to make antitrust law but an accurate description of where things stand.
Pull the current filings here: United States v. RealPage. Because the landlord cases and the private class actions move on separate schedules, case alerts are the practical way to keep all of them in one view.
The states' jury verdict against Live Nation in April 2026 is the closest recent example of a monopolization theory actually tested in front of a jury, covered in the Live Nation and Ticketmaster litigation. For a monopolization case now in its remedy and appeal phase, see the Google search monopoly appeal. And for the FTC's parallel effort against a platform's pricing and seller conduct, read FTC v. Amazon.