What the RealPage settlement requires
On November 24, 2025, the Justice Department announced a proposed settlement with RealPage, which needed court approval at the time. Under its terms, RealPage would:1
- Stop having its software use competitors' nonpublic, competitively sensitive information when setting rents.
- Stop using active lease data to train its models, and limit training to historic nonpublic data aged at least 12 months.
- Not use models that determine geographic effects at narrower than a state level.
The settlement binds RealPage, not every vendor. But it shows what federal enforcers are concerned about, and it's a useful yardstick for any tool you use.
What New York's law does
New York's S.7882 goes further. Morgan Lewis summarizes it this way: it's unlawful to provide a tool with a "coordinating function" to residential owners or managers, or to set rents or lease terms based on one's recommendations. A tool does this if it collects rent, supply or lease data from two or more unaffiliated owners or managers, analyzes it, and recommends rents, renewals, occupancy targets or other terms.2
Three points are worth noting for owners:
- It applies to owners and managers who use the tool, not just vendors.
- Liability covers acting "knowingly or with reckless disregard."
- It reaches multi-landlord data even when that data is public.2 That goes well beyond the RealPage settlement's focus on nonpublic data.
Other places to check
Morgan Lewis also lists San Francisco, Seattle, Jersey City, Minneapolis, Philadelphia and San Diego among places that have passed related laws, and notes that Berkeley's enforcement is paused during litigation.2 Other jurisdictions may have acted since. Check every market you own in, and check state law too.
What's lower risk, and what's higher
| Practice | General direction | Why |
|---|---|---|
| Pricing from your own leases and demand | Lower | No competitor data involved |
| Manually reviewing public competitor listings | Lower in most places | But note New York's approach to automated tools |
| Tool ingests public data from many owners and recommends rents | Check closely | May fall within laws like New York's |
| Tool pools competitors' nonpublic rents to recommend yours | Higher | The core concern in the RealPage matter |
| Staff told or pressured to accept recommendations | Higher | Looks more like coordination |
Ten questions for your manager or vendor
- What data trains the model, and where does it come from?
- Does it include any nonpublic data from other owners or managers?
- Does it use data from unaffiliated owners, including public data?
- How old is that data, and how is it aged?
- Is my data used to make recommendations for anyone else?
- At what geographic level does the model work?
- Can staff freely override a recommendation? How often do they?
- Are there auto-accept features or pressures to accept?
- How do you handle markets with special laws, such as New York?
- Will you put these answers in writing, and indemnify us?
Good practice for owners
- Inventory every pricing tool across the portfolio, including those your managers use.
- Have counsel review the tools, market by market.
- Keep records showing that people make the pricing decisions.
- Review the management agreement: who bears the risk if a manager's tool breaks the law?
- Avoid discussing rents or occupancy strategy with competitors, whether at trade events or anywhere else.
How this applies to us
We use AI in pricing, and we expect owners to ask us the questions above. Bring them to a working session. For wider context, see AI property management and AI vs traditional management.
