In 2022, a ProPublica investigation and subsequent legal actions brought the RealPage algorithmic pricing story to public attention. The allegations: that RealPage's YieldStar software facilitated coordinated rent-setting among competing landlords by using competitor data to recommend pricing that benefited the industry at the expense of renters.
Whether every specific allegation is true is for the courts to determine. But the story crystallized a genuine question that has become more urgent as AI enters the property management software market: what is AI in real estate actually for? Whose interests does it serve? And what does the responsible version look like?
These are questions Freehold has thought hard about—because we believe the answer matters, both for renters and for the independent landlords who are the backbone of the housing supply in most American cities.
What the RealPage Case Was Actually About
The core of the RealPage controversy wasn't that AI was used to set rents. It's that the alleged use of AI enabled coordination among competing property owners by aggregating competitor data in ways that, critics argued, resulted in above-market rents across entire markets.
Antitrust law has historically focused on explicit price-fixing agreements between competitors. The RealPage allegations introduced a new question: can algorithmic coordination achieve the same effect as explicit collusion, even if no human beings sat in a room and agreed to raise prices together?
Several major cities and the Department of Justice pursued this question in court. Regardless of the ultimate legal outcome, the commercial real estate software industry has been forced to reckon with it.
For independent landlords, the relevant lesson is different: the RealPage story is about large institutional operators coordinating through shared software. Most independent landlords manage 10–200 units, are not coordinating with anyone, and are often the ones being priced out of competitive dynamics that concentrated ownership and algorithmic pricing create.
The Real Problem with Institutionally-Oriented AI
RealPage's product was built for institutional operators—REITs, large management companies, entities with hundreds or thousands of units. The intelligence it delivered was designed to maximize revenue at that scale, in competitive markets where the operator has significant market presence.
For independent landlords, applying that same framework creates several problems:
Optimizing for market share doesn't make sense at 40 units. An operator with 3,000 units in a single market has enough presence to meaningfully influence local rent dynamics. An operator with 40 scattered units does not. Pricing strategies designed for market-level optimization don't translate to the independent landlord's actual situation.
Coordination-oriented pricing is bad for the market. Even apart from the legal questions, pricing software that primarily works by referencing what other landlords are charging creates an incentive structure that can drift toward coordination. For an independent landlord, that's both a legal exposure and an ethical one.
It doesn't tell you what you actually need to know. The independent landlord's pricing problem is not "how do I maximize rent extraction in a competitive market?" It's "am I charging fair market value for this specific unit, and am I leaving money on the table?" These are different questions that require different data.
What Ethical AI for Landlords Actually Looks Like
There's nothing inherently problematic about using data and software to help landlords make better pricing and management decisions. The question is what kind of intelligence you're building and for whom.
Here is the framework that Freehold is built around:
Owner Intelligence, Not Market Coordination
The goal is to help each independent landlord understand their own portfolio better—not to help a set of landlords coordinate with each other. Freehold benchmarks your rents against public market data (active listings, closed lease comps) to help you understand where you are relative to the market. It doesn't aggregate your data with other Freehold users' pricing to inform recommendations. Your competitors aren't in your dataset.
This is a fundamental architectural choice. An AI assistant that tells you "comparable units in your submarket are renting at $1,150–$1,250, and your unit is at $1,050" is giving you market information. An AI assistant that tells you "based on your competitors' pricing behavior in our network, you should raise to $1,175" is facilitating coordination. Only one of those is acceptable.
Surface Decisions, Not Just Metrics
A metric tells you what happened. A decision surfaces what you should do about it. The RealPage model provided pricing recommendations that optimized a single metric (revenue per unit) without surfacing the full decision context: tenant retention risk, local competitive dynamics, the difference between a unit that should command a premium and one that doesn't.
Ethical AI for landlords surfaces complete decisions: "Your two-bedroom on the second floor is 14% below market. At renewal, you could raise by $175 without exceeding comparable units. Your tenant has been in place 3 years with no payment issues—turnover here would likely cost $5,500–$7,000. A full market correction this cycle carries meaningful vacancy risk; consider a 7–9% increase as a retention-conscious path to market."
That's useful intelligence. It helps you make a better decision. It doesn't make the decision for you—and it doesn't make it by coordinating with your neighbors.
Transparency in Methodology
Landlords should understand how any AI-driven recommendation is generated. What data is it drawing on? How is it benchmarking? What assumptions is it making? Black-box recommendations from large software vendors are a feature of the institutional market, where buyers have compliance and legal teams to evaluate them. Independent landlords don't. They need to understand what they're buying.
Freehold's methodology is documented and explainable. Every recommendation comes with the underlying data and logic visible—not buried in an algorithm you're asked to trust.
Designed for the Tenant Relationship, Not Against It
The independent landlord and their tenants are in a long-term relationship. The average tenancy in a well-managed small multifamily building is 2–4 years. Turnover is expensive. Stability is valuable.
AI that optimizes purely for rent maximization without accounting for tenant relationship quality is optimizing a single variable in a multi-variable system. The result, as the RealPage story suggests, can be rent increases that generate short-term revenue at the cost of tenant stability, community fabric, and the landlord's long-term operating performance.
Freehold's rent intelligence includes tenant retention context as a feature, not an afterthought.
The Housing Market Context
It would be incomplete to write this article without acknowledging the broader housing affordability crisis. In many American cities, rents have increased substantially over the past five years, housing supply has not kept pace with demand, and the financial pressure on renters is real and significant.
Independent landlords are not the cause of this crisis—they are often small operators providing housing in markets that are structurally supply-constrained. But they operate in a policy environment that is increasingly scrutinizing algorithmic rent-setting, and they should understand what that scrutiny is about.
The right position for an independent landlord is to price at market value—not below it (which leaves money on the table and may create perverse vacancy dynamics), not above it (which creates vacancy and, at scale, contributes to market distortion), and not through coordination with competitors. Market rate, for your specific unit, based on transparent public data.
That's not just ethically correct. It's the operationally rational position.
Freehold's Commitment
Freehold is building AI for independent landlords—not for institutions. The product is designed to close the information gap between what the institutional market knows and what the independent operator knows, without adopting the institutional market's incentive structures.
That means: no competitor data aggregation, no market coordination mechanics, full transparency in methodology, and intelligence designed to make each individual owner smarter about their own portfolio—not to make the landlord class collectively richer at the expense of renters.
We think that's the version of AI that independent landlords should be looking for. And we think the RealPage story is a useful reminder of what happens when the design choices go the other way.
Bottom Line
The RealPage controversy is a story about incentive design: software built to serve institutional coordination rather than individual owner intelligence. The problems that emerged were predictable from the design.
Ethical AI for landlords looks different. It uses public market data to help each owner understand their specific situation. It surfaces decisions with full context, not single-variable optimizations. It's transparent about methodology. And it treats the landlord-tenant relationship as a long-term asset to be managed well, not a revenue-maximization opportunity to be exploited.
That's not a harder product to build. It's a different set of priorities. And for the independent landlord who wants to run a better portfolio without becoming a case study in what AI for real estate shouldn't be—it matters.