Key takeaways
- The DOJ and RealPage agreed to settle one of the most high-profile ongoing antitrust cases, signaling a continued enforcement focus on combatting algorithmic coordination and improper information sharing.
- The proposed settlement targets RealPage’s use of real time nonpublic data by limiting the data RealPage is permitted to use to train its software models to historical data at least 12 months old. The proposed settlement further restricts the granularity with which RealPage can report rental pricing information with its subscribers by forbidding RealPage from providing data more narrowly than at a state-wide level.
- The proposed settlement, which includes no financial penalties, or findings of wrongdoing, requires court approval.
- Companies should be mindful of the models and data used in any algorithmic pricing tools they employ and should consult counsel to ensure compliance with antitrust laws.
In depth
On November 24, 2025, the DOJ's Antitrust Division announced a proposed settlement to resolve its claims against RealPage Inc. This action is part of a broader federal enforcement effort to confront increasing use of algorithmic tools to engage in coordination, share competitively sensitive information, and pursue other potentially anticompetitive practices, including in the rental housing industry.
According to allegations in the DOJ's complaint, RealPage's revenue management software used nonpublic, competitively sensitive information shared by landlords to set rental prices. The Complaint alleged violations of Section 1 and 2 of the Sherman Act through agreements to share and use competitively sensitive information to align user landlords’ pricing processes, strategies, and pricing responses via the software. The DOJ alleged that this practice undermined independent pricing decisions among competing companies and allowed for alignment of rental prices, harming both competition and consumer choice.
The proposed consent judgment details several corrective measures that RealPage must undertake if the settlement is approved by the court.
|
Key Settlement Provisions
- Real-time data: RealPage cannot use active lease data to train its algorithms--only historical or backward-looking data aged at least 12 months.
- Geographic models: RealPage's pricing models may not analyze geographies narrower than a state level.
- Feature redesign: RealPage may not make identical pricing recommendations to different owners within the same geographic market, and must modify its software features to avoid limiting price decreases or aligning pricing among competing landlords.
- Market intelligence: RealPage must not conduct, or solicit nonpublic data through market surveys for purposes of recommending pricing. Nor may they use, share, or disclose current, forward-looking data, nor the data collected through market surveys, for the use of the software.
- Information exchange: RealPage must refrain from discussing or facilitating discussions on market analyses or trends based on nonpublic data, or pricing strategies, in meetings relating to the software.
- Oversight: A court-appointed monitor will oversee RealPage’s compliance.
- No financial penalties or admission of wrongdoing: No fines, or admissions of liability against RealPage.
|
The terms of the settlement will be effective for seven years from its entry date, but the DOJ may end it after four years if the DOJ no longer finds it necessary or in the public interest.
RealPage and its landlord clients have also faced class action litigation and separate lawsuits from attorneys general from several states, including Maryland, New Jersey, Kentucky, Washington, Arizona and the District of Columbia. Some of those actions also have settlements pending.
Conclusion
The RealPage settlement demonstrates that there are significant antitrust risks in using AI-based pricing tools where algorithms rely on competitors' nonpublic, current data in ways that may align pricing or impact independent decision-making. Companies that subscribe to or market such pricing models should ensure they rely on only backward-looking, aged data, avoid the ingestion of competitors' real-time pricing data, and eliminate any features that limit or discourage price decreases. Such pricing algorithms should also avoid overly granular geographic modelling that could facilitate local coordination.
Companies should consult with antitrust counsel prior to using or marketing AI pricing tools and with respect to information exchange generally.
