The AI landlord story has moved from lawsuit to settlement, and the numbers involved are now large enough to reshape how the rental industry uses pricing software going forward. On 24 November 2025, the US Department of Justice reached a proposed settlement with RealPage, the company whose algorithmic pricing software sits at the centre of the DOJ’s original August 2024 antitrust lawsuit accusing the firm of enabling landlords to coordinate rents through a shared pricing algorithm rather than competing independently.
The DOJ’s original complaint, filed alongside eight state attorneys general, alleged that RealPage’s software let landlords feed nonpublic, competitively sensitive lease data into a shared algorithm that then recommended rents back to all of them, a practice Attorney General Merrick Garland described at the time as landlords finding “a new way to scheme” with software standing in for a phone call. RealPage has not admitted wrongdoing in the settlement, but agreed to limit how it uses active lease data to train the predictive models behind its Revenue Management Solutions, Lease Rent Options and AI Revenue Management products.
The Money Behind the AI Landlord Lawsuits
Separately from the DOJ settlement, a private class-action lawsuit consolidated in the US District Court for the Middle District of Tennessee has produced a much larger financial reckoning. Fourteen housing companies agreed in May 2026 to pay a combined $218 million to resolve claims that they conspired with rivals to inflate rents using the same AI landlord software, following an earlier October 2025 batch of 26 settlements worth more than $141.8 million that included a $50 million payment from Greystar, the largest US apartment landlord.
Equity Residential’s $56 million payment is now the largest single settlement in the case, with Camden Property Trust and Mid-America Apartment Communities each paying $53 million. The combined settlement total across both rounds has reached nearly $360 million, a figure that reflects how many separate landlords adopted the same AI landlord pricing tool rather than any single company acting alone.
Every settling landlord also agreed to stop feeding RealPage their own nonpublic lease data and to stop using any version of the software that draws on competitors’ nonpublic pricing information, the specific mechanism the DOJ and private plaintiffs both argued turned individual pricing decisions into coordinated ones. RealPage itself, still a defendant in the ongoing litigation, has denied wrongdoing throughout and maintains that no changes are needed to the products its remaining customers use.
Tenant screening software used alongside AI landlord pricing tools raises a related but distinct concern LiveAIWire has traced in its reporting on predictive risk tools in criminal justice: proxy variables like ZIP code or credit history gaps can encode historical inequality into a screening score without the software ever referencing a protected characteristic directly, a pattern that shows up in AI landlord applicant screening as readily as it does in courtroom risk assessment.
RealPage Is Now Suing to Stop Cities From Banning the AI Landlord Model
Rather than retreating after the settlements, RealPage has gone on the offensive against the local governments trying to restrict algorithmic rent-setting outright. The company is suing New York State over a new law banning algorithmic rent-setting software, arguing in its complaint that the version of its software now permitted under the DOJ’s own consent decree does not reference any competitor’s nonpublic information and therefore cannot plausibly be used to facilitate collusion. Critics were quick to note the tension in that argument: RealPage is citing a settlement it entered while denying liability as evidence that its product should now be treated as legally safe.
New York is not an isolated case. Bellingham, Washington is considering becoming the latest city to ban rent-pricing software outright, following a wave of similar local measures since 2024. Nine states reached a separate $7 million settlement with Greystar specifically over its use of RealPage software, under which Greystar agreed to stop using any rent-setting tool that relies on other landlords’ private data, a condition now echoed across most of the major settlements in the federal case.
What the Settlements Actually Change, and What They Don’t
The practical effect of the settlements so far is narrower than a ban on AI landlord software itself. What is now restricted, across the DOJ consent decree and the private settlements alike, is the specific practice of pooling competitors’ nonpublic lease and pricing data into a shared model. Pricing software built on a landlord’s own historical data, without ingesting rivals’ current, non-public figures, remains squarely permitted, which is exactly the distinction RealPage is now relying on in its lawsuit against New York’s broader ban.
That narrower scope is precisely why some cities are choosing to legislate beyond it rather than wait for further federal enforcement. A settlement that restricts data-sharing addresses the antitrust theory the DOJ pursued, coordination between competitors, but it does not address the separate concern that a single landlord’s own AI landlord pricing tool might still set rents higher than a human property manager would have, simply because the software is optimised purely for revenue rather than for retention, tenant relationships, or any of the softer considerations a human decision-maker might weigh.
That distinction matters for tenants because it determines how much actually changes in practice. A rent-setting algorithm trained only on a landlord’s own portfolio history can still recommend aggressive pricing; what it can no longer legally do, under the settlements reached so far, is factor in what a competing landlord’s software is recommending for a comparable unit down the street in real time. The underlying tension, between an algorithm that is individually rational and one that produces collectively coordinated outcomes, echoes LiveAIWire’s coverage of the AI credit score’s persistent lending gap, where algorithmic decision-making narrowed one kind of discrimination while leaving a different, harder-to-see pattern largely intact.
What This Means for Renters
For a tenant trying to understand why a renewal offer jumped sharply, the settlements provide a partial but incomplete answer. If a landlord’s pricing software is still using only that landlord’s own data, a steep increase may reflect genuine market conditions rather than coordinated pricing, and the settlements do nothing to cap that. If a landlord has not yet implemented the required changes, or is using a different vendor’s software built on similar principles, the same coordination concerns that drove the original lawsuits could still apply.
Tenants in cities actively considering their own bans, including Bellingham, have more direct leverage than tenants elsewhere, since local ordinances can restrict the practice outright rather than relying on the narrower data-sharing limits the federal settlements impose. The broader dynamic, in which the economic gains from a new technology concentrate with the companies deploying it while the costs land on people with the least power to contest a pricing decision, is one LiveAIWire has traced in its coverage of the wider AI automation divide.
Whether more cities follow San Francisco, Philadelphia, Minneapolis and now potentially Bellingham in passing outright restrictions on AI landlord software is likely to be shaped as much by how RealPage’s lawsuit against New York’s law resolves as by any single city’s own legislative process. A ruling in RealPage’s favour would give every AI landlord vendor a template for arguing that removing competitor data satisfies the law, while a loss would strengthen the case for the kind of outright bans Bellingham and New York are now considering.
Stuart Kerr is Technology Correspondent at LiveAIWire, covering artificial intelligence, emerging technology, and their impact on business, society, and everyday life.