She had been renting the same apartment for three years. The neighborhood had not changed. The building had not been renovated. Her landlord had not mentioned anything about costs going up. Then the lease renewal arrived, and the rent had jumped $200 a month. When she asked why, the property manager said it was market rates. That was true, technically. What was also true what she did not know, what most renters did not know, that her landlord had not decided that number independently. Software had.
The software was called RealPage. And it was running in apartment buildings across 43 states.
What the Algorithm Actually Did
RealPage sold itself as a revenue management tool for landlords, a way to optimize rental pricing using data and market intelligence. What the Department of Justice alleged, in a lawsuit filed in August 2024 and expanded in January 2025, was something more specific and more damaging. The DOJ complaint accused RealPage of collecting non-public, competitively sensitive pricing data from competing landlords, feeding it into an algorithm, and using it to generate rental price recommendations that eliminated the downward pressure competition normally creates.
In plain terms: landlords who were supposed to be competing with each other for tenants were all looking at the same software, which had been trained on everyone else’s private lease data, and following its recommendations in sync. That is not a market. That is coordination. The Sherman Antitrust Act, passed in 1890, exists precisely to prevent it.
The DOJ expanded its lawsuit in January 2025 to include six of the largest property management companies in the country: Greystar Real Estate Partners, Blackstone’s LivCor, Cushman & Wakefield, Cortland Management, Camden Property Trust, and Mid-America Apartment Communities. Together, these landlords managed more than 1.3 million rental units across the country. They were not small operators making independent decisions. They were institutional players using the same tool to arrive at the same prices, in city after city, without ever having to pick up a phone.
The Number That Explains the Rage
Renters paid an estimated 5 to 7 percent more than they would have in a competitive market. That is the figure that emerges from multiple analyses of the RealPage case. On a $1,500 apartment, that is $75 to $105 a month. On a $2,000 apartment, it is $100 to $140. Every month. For years.
The algorithm had been operating since at least 2016. The DOJ lawsuit covered conduct through 2024. Eight years of above-market rents, in buildings across 43 states, affecting millions of tenants who had no idea the pricing they faced had been coordinated rather than competed. The property manager who told the renter it was “market rates” was not lying. The market rate had been engineered.
What RealPage’s software did, specifically, was host a system where participating landlords shared forward-looking pricing data and lease terms information that competitors in any functioning market would keep private. That data trained the algorithm. The algorithm generated recommendations. The landlords followed them. The DOJ alleged that RealPage also hosted “user group” meetings where competing property managers discussed how to modify the software’s pricing methodology, and that managers engaged in “call-arounds,” direct conversations where they shared pricing information with competitors under the guise of market surveys.
This was not passive software running in the background. It was an active coordination mechanism with a user interface and a meeting schedule.
The Firms Behind the Algorithm
Greystar Real Estate Partners manages more than 800,000 apartment units globally, making it one of the largest residential landlords in the world. Blackstone, the same firm that has spent billions acquiring assets across American healthcare, manufactured housing, and commercial real estate, operated its multifamily residential portfolio through LivCor, named as a co-defendant. Cushman & Wakefield is one of the largest commercial real estate services firms on earth.
These are not corner landlords raising rent because heating costs went up. These are trillion-dollar institutional operators with analytics teams, revenue optimization departments, and shareholder obligations to maximize returns on residential real estate. RealPage gave them a mechanism to do that collectively, without formally agreeing to do so, or at least, that was the allegation. The legal distinction between an algorithm that coordinates pricing and a handshake agreement to fix prices is what the lawsuit was designed to test.
RealPage defended itself by arguing that its software was used by less than 10 percent of US rental units and that housing shortages, not algorithmic coordination, were the primary driver of high rents. Both things can be true simultaneously. A shortage creates upward pressure. An algorithm that eliminates competitive downward pressure ensures that pressure is never relieved.
The Settlement That Did Not Admit Anything
On November 24, 2025, the DOJ filed a proposed settlement resolving its antitrust claims against RealPage. RealPage did not admit liability. What it agreed to do was significant regardless: stop using competitors’ non-public data in its revenue management product, limit algorithm training to historical data at least 12 months old, and submit to three years of monitoring by the DOJ.
The settlement covered only the federal claims. State attorneys general in multiple states, including New York, which had passed new legislation explicitly targeting algorithmic rent pricing, did not settle. Private class action lawsuits from renters remain active. Nevada reached early settlements with RealPage and several landlords totaling more than $140 million. That money is real, but spread across millions of affected tenants, it amounts to a fraction of the excess rent they paid over eight years.
Under the settlement terms, RealPage can no longer use real-time competitor data to generate price recommendations. The software still exists. The company still operates. The landlords who used it are still managing their portfolios. What changed is one data input. The infrastructure that made coordinated pricing possible remains largely intact, now governed by a consent decree that expires in three years.
The Steelman: What RealPage’s Defenders Got Right
The strongest defense of algorithmic pricing tools is also the least comfortable to dismiss.
Housing shortages are real. In markets where demand consistently outpaces supply, rents will rise regardless of what software landlords use. The cities where RealPage’s impact was alleged to be most significant, Austin, Atlanta, Phoenix, and Denver, are also cities where construction has lagged behind population growth for decades. Blaming an algorithm for price increases driven by fundamental supply constraints gives renters a satisfying villain without solving the underlying problem.
There is also a legitimate argument that revenue management software, used transparently, can improve market efficiency. A landlord who prices a unit based on actual local market data, publicly available comparable listings, neighborhood trends, and seasonal patterns is doing something different from a landlord using competitors’ private forward-looking lease agreements. RealPage crossed a line that the DOJ could identify and prosecute. Other pricing tools using only public data remain legal and widespread.
The defense ends when you read the DOJ complaint’s description of the “call-around” meetings, the user groups where competitors discussed modifying pricing methodology together, and the algorithm’s explicit design to reduce the frequency and magnitude of rent decreases. A tool built to prevent rents from falling, trained on private competitor data, is not a market efficiency tool. It is a price floor mechanism. The DOJ’s case rested on that distinction, and the settlement terms suggest the government found enough to make it hold.
What Every Renter Should Understand
The RealPage case is not an anomaly. It is the most documented example of a broader shift in how residential real estate is managed in the United States.
Institutional landlord firms managing thousands or tens of thousands of units operate differently from individual property owners. They have data science teams. They use yield management software borrowed from the airline and hotel industries. They optimize occupancy rates and revenue per unit across portfolios that span multiple cities simultaneously. The renter sitting across the lease table is not negotiating with a person who owns a building and needs to pay a mortgage. They are negotiating with a revenue optimization system that has already calculated the maximum price the local market will bear.
The same household budget pressure that is feeding broader economic fragility across 2026 flows directly through rent. It is the single largest expense for most American households. When that expense is set not by competition but by coordinated algorithmic recommendation, the mechanism that is supposed to keep prices honest, the threat that a competing landlord will offer a lower price is neutralized. The tenant has nowhere cheaper to go, because the cheaper option is using the same software.
New York and California have both passed legislation targeting algorithmic rent pricing. The EU is watching. The DOJ case established that the behavior is prosecutable. None of that restores the money renters paid between 2016 and 2024, or changes the fact that the same structural forces pushing homeownership out of reach for an entire generation also ensured that renting was made more expensive through coordination rather than competition.
The Market Rate Was Not the Market
The property manager who said it was market rates was not wrong. The market rate in 2022 in that building, in that city, was whatever RealPage said it was. It was market rates in the same way that a price fixed by a cartel is technically accurate, structurally dishonest.
The renter who asked why her rent went up got an answer that ended the conversation. What she did not get was the part that came before the number, the algorithm, the user groups, the shared competitor data, the DOJ lawsuit, the eight years of coordinated recommendations that made “market rates” mean something different from what the words suggest.
She paid the $200 increase. Most people did. The software had calculated, correctly, that they would.
