SB52 makes it unlawful to sell, license, provide, or use so‑called rental pricing algorithms when those tools are intended or reasonably expected to be used by multiple landlords in the same or related rental market to set or recommend rental terms. The bill targets algorithms that process "nonpublic competitor data"—information about actual rents, occupancy, and lease dates obtained through nonpublic means—and treats distribution, use, and each affected unit and month as distinct violations.
The law combines a definition regime (what counts as a rental pricing algorithm and what data are off‑limits), civil enforcement by the Attorney General and local prosecutors, and a private right of action with statutory penalties (up to $1,000 per violation) and fee shifting. For vendors, property managers, and landlords that rely on revenue‑management software, the bill imposes a new compliance architecture and litigation exposure; for tenants and local governments, it offers a legal tool to challenge algorithm‑driven price coordination.
At a Glance
What It Does
SB52 forbids selling, licensing, or providing rental‑pricing algorithms when those tools are intended for use by two or more landlords in the same or related market and it forbids using such algorithms when one knows (or should know) they will be used across landlords. It also bans adopting rental terms based on algorithms that process nonpublic competitor data already used by another landlord.
Who It Affects
Proptech vendors, revenue‑management and pricing‑software companies, institutional landlords and property management firms, and any landlord using automated pricing recommendations; tenants, city and county prosecutors, and the Attorney General are given enforcement standing.
Why It Matters
The bill explicitly folds algorithmic pricing into California’s antitrust posture and creates a statutory enforcement path separate from—but cumulative with—existing antitrust laws. That combination forces product redesign, compliance controls, and legal risk assessments across the rental‑pricing ecosystem.
More articles like this one.
A weekly email with all the latest developments on this topic.
What This Bill Actually Does
SB52 draws a bright line around certain automated pricing products: if a service uses nonpublic, competitor‑level data about actual rents, occupancy, or lease dates to advise multiple landlords in the same or related market about rents, lease lengths, or occupancy targets, the bill makes it unlawful to distribute or use that service. The law is careful to call out both the commercial side (selling, licensing, or providing the software) and the operational side (using it to set or adopt rental terms), and it adds an explicit anti‑coercion rule covering situations where one landlord or vendor pressures another to follow algorithmic recommendations.
Definitions drive compliance. "Nonpublic competitor data" is the linchpin: the bill prohibits using data pooled from two or more competitors obtained by nonpublic means, but it carves out a long list of exceptions (public listings, municipal rental registries, census data, aggregated non‑linkable reports, data more than one year old, and MLS‑style listings). It also excludes simple aggregated reports that publish public data without recommending future rates, and products used solely to set rents under affordable housing program rules.SB52 multiplies liability in several ways: each month a violation continues counts as a separate violation, each month a vendor provides the tool in violation counts separately, and each residential unit affected counts separately.
Enforcement is both public and private: the Attorney General and local prosecutors can sue for injunctions, restitution, and penalties up to $1,000 per violation, and harmed persons can bring private suits with the same remedies; prevailing plaintiffs recover fees, and lease clauses waiving tenant fee recovery are void as to these claims.Practically, vendors must reexamine how they collect or ingest data, how models are trained on competitor information, and how distribution agreements and APIs are structured to avoid an intent or reasonable expectation that the product be used by multiple market participants. Landlords and property managers must decide whether to stop using certain automated recommendations, document their data sources, and update vendor contracts.
Regulators and courts will confront several interpretive questions—particularly what constitutes a "related market," what it means to "know or should know" about cross‑user deployment, and how to apply the one‑year data rule in environments with frequent listings.
The Five Things You Need to Know
SB52 makes it unlawful to sell, license, or otherwise provide a rental pricing algorithm to two or more persons when the vendor intends or reasonably expects the tool will be used by multiple landlords in the same or related market.
The statute defines "nonpublic competitor data" as information about actual rents, occupancy rates, and lease start/end dates derived from two or more competitors obtained through nonpublic means, and explicitly excludes public listings, rental registries, census data, aggregated non‑linkable reports, and data older than one year.
The bill treats each month of use or distribution and each separate residential premises affected as a distinct violation, multiplying exposure for ongoing use or wide distribution.
Enforcement includes civil actions by the Attorney General and city or county prosecutors and a private right of action; penalties are up to $1,000 per violation, and prevailing plaintiffs recover attorney’s fees—tenant lease clauses that limit fee recovery are void for these claims.
A parent company and its wholly owned subsidiaries are treated as one person for the statute’s purposes, and the law excludes products that only publish aggregated public data or that establish rents under affordable‑housing program rules.
Section-by-Section Breakdown
Every bill we cover gets an analysis of its key sections.
Ban on selling or providing rental pricing algorithms for multi‑landlord use
Subdivision (a) targets vendors and intermediaries: it makes it unlawful to sell, license, or otherwise provide a rental pricing algorithm to two or more persons when the supplier intends or reasonably expects the tool will be used by multiple landlords in the same or a related market. For compliance teams this is a distribution‑control rule—how a product is positioned or how licensing terms are written may determine liability as much as how the software functions.
Prohibition on using algorithmic recommendations and anti‑coercion rule
These subsections bar a landlord's use of a rental pricing algorithm when the user knows or should know the algorithm would be used by multiple landlords, and they prohibit coercing another person to adopt recommended rental terms. The language blends a mens‑rea standard—"knows or should know"—with an explicit coercion prohibition, creating both strict and negligence‑style exposure depending on proof of what the user knew or ought to have known.
Multiplicity of violations: months, units, and distributions
Subdivision (d) converts ongoing or repeated acts into serial violations: each month a violation continues counts separately, each month a vendor provides the offending algorithm counts separately, and each distinct residential premises affected counts separately. That structure substantially increases potential statutory penalties and makes continuous monitoring and remediation important once a compliance failure is identified.
Key definitions and narrow exclusions for covered data and products
Subdivision (e) supplies the statute’s operative definitions: what constitutes antitrust laws, "nonpublic competitor data," "nonpublic data," and a "rental pricing algorithm." The bill lists multiple exclusions from nonpublic competitor data (public listings, rental registries, census data, aggregated non‑linkable reports, data older than one year, and MLS listings) and excludes products that merely publish aggregated public data or that set rents for affordable housing programs. Those carve‑outs will guide product design and data‑ingestion policies.
Enforcement, civil remedies, and fee shifting
Subdivision (f) authorizes the Attorney General and local prosecutors to bring civil actions for injunctions, restitution, and penalties up to $1,000 per violation and gives harmed persons a private right of action with identical remedies. The statute mandates fee shifting to prevailing plaintiffs (including public prosecutors) and voids lease provisions that bar or cap tenant recovery of attorney’s fees in actions under this section, increasing litigation leverage for plaintiffs and enforcement agencies.
Savings clause: cumulative with antitrust laws
Subdivision (g) clarifies that the new prohibitions do not replace existing state or federal antitrust laws but sit alongside them; remedies are cumulative. Practically, plaintiffs may bring parallel claims under this statute and antitrust laws, and courts will need to reconcile overlapping doctrines and damages frameworks.
This bill is one of many.
Codify tracks hundreds of bills on Housing across all five countries.
Explore Housing in Codify Search →Who Benefits and Who Bears the Cost
Every bill creates winners and losers. Here's who stands to gain and who bears the cost.
Who Benefits
- Renters and tenant advocacy groups — SB52 reduces one pathway for algorithmic rent convergence and coordinated price increases by restricting tools that pool nonpublic competitor data, giving tenants a statutory claim to challenge algorithm‑driven rent setting.
- Local governments and housing enforcement agencies — the bill provides city and county prosecutors explicit authority to sue for injunctions and penalties, supplementing existing enforcement tools against coordinated or anti‑competitive rental practices.
- Independent landlords who set prices locally — landlords who prefer hands‑on, market‑based pricing without relying on pooled competitor data may benefit from reduced upward pressure on rents caused by cross‑user algorithms.
Who Bears the Cost
- Proptech vendors and revenue‑management companies — firms that collect competitor rent and occupancy data or license pricing recommendations to multiple landlords will need to redesign data ingestion, model training, and licensing practices or risk serial statutory violations.
- Large institutional landlords and property managers — organizations that rely on automated pricing at scale face compliance costs, potential structural changes to pricing workflows, and exposure to monthly, per‑unit penalties where algorithms used by their vendors violate the statute.
- Legal and compliance departments — both vendors and landlords must invest in contract revisions, audit trails proving data sources and distribution intent, and potentially defend against private and public enforcement actions that could be frequent given the statute’s fee‑shifting and per‑violation structure.
Key Issues
The Core Tension
The central trade‑off is between preventing algorithmic collusion that inflates rents and preserving the legitimate efficiency benefits of data‑driven pricing: the bill protects competition and tenants but risks chilling product innovation and imposing heavy compliance and litigation costs on vendors and landlords that use automated pricing tools responsibly.
SB52 is precise in some respects but leaves several implementation questions that will drive litigation and regulatory guidance. The statute hinges on phrases such as "same market or a related market," "knows or should know," and "nonpublic competitor data." Each of those terms requires factual inquiry: markets for rental housing can be hyperlocal or regional depending on the property type, and vendors may intentionally design products to be useful across markets without targeting coordination.
Proving what a vendor or landlord "should know" about cross‑user deployment will likely turn on procurement documents, licensing language, and internal communications.
The bill’s list of exclusions mitigates overbreadth but invites workarounds and interpretation battles. The one‑year lookback exclusion for data processing will pressure vendors to retain historical data outside the one‑year window or argue that older data are permissible for training models.
Aggregated, non‑linkable reports are exempt, but the line between an aggregated market report and an actionable recommendation can be thin. The multiplicative penalty structure—per month, per unit, and per distribution—creates large potential exposure for recurring conduct and heightens the risk of nuisance litigation given the $1,000 cap is modest per instance but can balloon in aggregate; fee shifting further increases plaintiff leverage.
Finally, treating parents and wholly owned subsidiaries as one person reduces corporate arbitrage but complicates cases where ownership structures are less than full control.
Try it yourself.
Ask a question in plain English, or pick a topic below. Results in seconds.