by Samuel Leitch
In 2022, a time when rents continued to soar impossibly, the company RealPage boasted that it could help landlords increase profits even further.1 How was this possible? By pooling nonpublic pricing data from clients, RealPage’s software can offer landlords recommendations on the highest possible rent that they can set for a given residential unit. Although it would be illegal for these landlords to communicate directly to agree upon rent prices, critics of RealPage claim that this software simply acts as an algorithmic middleman for price setting. RealPage is no small player, either: Greystar, the largest apartment manager in the United States, will have to pay $7 million to nine different states for its use of RealPage’s software.2
This scenario is an example of algorithmic price fixing, where through the use of software, competitors may coordinate outcomes that would otherwise be illegal under antitrust law.
Existing antitrust laws in California, such as the Cartwright Act, prevent two or more people from collaborating to restrict trade. However, when it comes to algorithmic price fixing, a gray area arises. If two landlords do not directly agree to keep their rents high—for example, if instead they lend their data to an algorithm that leads to the same outcome, the same increase in rent—has collusion occurred?
Several bills have been introduced in the California state legislature this year with the goal of banning algorithmic price fixing. They were supported by a collection of civil society and labor groups, facing opposition from industry associations.
AB 325 (Aguiar-Curry): “Cartwright Act: violations”
Summary
AB 325 specifies that algorithmic price fixing is a violation under the Cartwright Act. It bans the use of pricing algorithms trained on nonpublic competitor data, and it also bans the use or sharing of any pricing algorithm so long as two or more companies use that algorithm to set or suggest prices.
Algorithmic price fixing is less direct than traditional agreements between companies. For this reason, it is harder to report. AB 325 makes it sufficient for complaints under the Cartwright Act to demonstrate that a trust or contract could have plausibly occurred, lessening the burden of proof.
AB 325 was passed by the Legislature and signed into law.
Opposition
Much like the bills this cycle that targeted surveillance pricing, AB 325 faced opposition from a coalition of industry partners led by the California Chamber of Commerce. Many of the opposition’s arguments were similar in both cases.
- The Chamber of Commerce wrote that this bill would stifle efficiency. As a result, profits would fall, customers would pay more, and small businesses in particular would suffer.
- Much like they did concerning surveillance pricing, the Chamber of Commerce took issue with the idea that pricing algorithms are necessarily predatory, arguing that in many cases these algorithms can lower costs for consumers. This, however, was a much smaller part of the arguments against AB 325 than with those against surveillance pricing.
- Another argument in common with those against surveillance pricing was to argue that existing law is already more than adequate in terms of preventing potential violations. In the words of the Chamber of Commerce, “collusion is collusion and is already effectively covered by existing law.”3
- It is worth noting that the California Attorney General has not brought a criminal prosecution under the Cartwright Act in decades, and in 2022, the California Law Revision Commission was tasked with modernizing California antitrust law.4 This suggests that perhaps antitrust law in California has not caught up with current advances in technology. Furthermore, the California Attorney General’s Office supported the bill, writing that “price fixing is illegal under existing law, but AB 325 simply makes it clear that using common pricing algorithms to fix prices among competitors is just as illegal as traditional price fixing.”5
- The opposition alleged that the bill’s language was overly broad; as a result, it would apply not only to algorithmic price fixing but also to all algorithms and even public data, having an effect so broad that business would be brought back to “pre-technological times.”6 The opposition made references to the many algorithms utilized by industry partners, perhaps to imply that these systems are too delicate for the broad brush of the legislator. For example, the Chamber of Commerce wrote that dynamic pricing, the adjustment of prices according to real-time demand, is one of many common pricing algorithms that would inadvertently fall under the bill’s purview. However, if a business conducts dynamic pricing in response to its competitor’s prices in real time, is the difference between that and algorithmic price fixing merely semantic?
SB 295 (Hurtado): “California Preventing Algorithmic Collusion Act of 2025”
Summary
SB 295 would have included both supply-side and direct-use provisions, targeting both the developers and vendors who make algorithmic pricing recommendations as well as the end users. The bill would have provided an affirmative defense so long as defendants could prove that they exercised due diligence: for example, by asking for confirmation from the developers that their algorithm does not use competitor data. Otherwise, each violation would incur a civil penalty of up to $25,000. Exceptions were also made for credit scores, which would not have been covered by the bill.
SB 295 failed passage on the California Assembly Floor after the third reading and was not signed into law.
Opposition
Many of the arguments against AB 325 were reused against SB 295, and certain paragraphs were copied-and-pasted between the letters of opposition against both bills.
- Again, the Chamber of Commerce took issue with the lack of distinction between pricing algorithms and collusion. In a letter of opposition, they wrote that “simply using a pricing algorithm does not evidence collusion or inherently constitute price fixing.”7 Similarly, they took issue with what they referred to as broad language, writing that this bill might unfairly target other helpful algorithms.
- The opposition argued that small businesses would be hurt most of all under SB 295.
- Like with AB 325, the opposition argued that since this bill will hinder efficiency, it will in turn raise costs for consumers. They wrote that pricing algorithms are common “tools that enable businesses to enhance efficiencies by avoiding manual pricing, saving money and lowering costs for consumers.”8
SB 52 (Pérez): “Housing rental terms: algorithmic devices”
Summary
SB 52 was similar to SB 295, the main difference being that SB 52 focused on algorithmic rent setting, capping penalties per violation at $1,000. It was supported by a wide coalition of civil rights and housing rights groups. The opposition consisted of realtors’ associations as well as RealPage.
SB 52 was held in committee and not signed into law.
Opposition
- For most of its lifetime, SB 52 included a ban on any usage of a rental pricing algorithm by any person, given that this person either knew that other landlords would use the algorithm or coerced other landlords into doing so. Although a later amendment softened this broad prohibition—instead of it being illegal to use a rental pricing algorithm, it would have been illegal to set rents based on the algorithm’s recommendations—this previous, much broader language was a major focus of the opposition’s arguments. The California Apartments Association argued that this language would hold “landlords to a much higher standard than users of other pricing algorithms under existing law.”9
- Using a similar playbook as the Chamber of Commerce, RealPage argued that SB 52 is too broad, potentially applying to “tools or products that analyze data without influencing rental pricing” and even being “interpreted to ban the use of all external data—both public and nonpublic.”10
SB 384 (Wahab): “Preventing Algorithmic Price Fixing Act: prohibition on certain price-setting algorithm uses”
Summary
Like some of the bills above, SB 384 had both supply-side and direct-use provisions. It would have prohibited both the sale and the usage of a price-setting algorithm in cases where two or more competitors would have used it to set prices or rents. It also would have provided an affirmative defense for users of these algorithms similar to the one that SB 295 would have provided. Lastly, SB 384 would have capped civil penalties at $1,000 per offense.
SB 384 was held in committee and did not become law.
Opposition
- As did opponents of the bills above, opponents of SB 384 wrote that it would harm small businesses and decrease competition. Similarly, opponents drew a distinction between pricing algorithms, which they claimed often benefit consumers, and price-fixing practices, which they maintained comprise a narrow subset of algorithms in practice. The strategy went as follows: first, to emphasize how extremely common it is for companies to use pricing algorithms, and second, to claim that existing law already effectively covers collusion. This approach frames legislation like SB 384 as an overreaction that risks destroying the delicate machinery of industry and its pricing tools.
- A critique more specific to SB 384 had to do with nonpublic data. In a letter of opposition, a coalition against SB 384 wrote that the “distinction between publicly available and nonpublic or confidential information is significant, because it preserves activities that businesses long performed in making pricing decisions, and done so legally,” such as to “…observe, analyze, and respond to market conditions; collect information on prices…and create pricing models to inform pricing decisions.”11 The distinction here is between price fixing and a range of practices that include dynamic pricing, adjustments based on local demand, and so on. A later version of SB 384 narrowed its definition of “nonpublic input data,” exempting data that was collected more than one year before the use of the algorithm.
1 Heather Vogell, “Rent Going Up? One Company’s Algorithm Could Be Why.,” ProPublica, October 15, 2022, https://www.propublica.org/article/yieldstar-rent-increase-realpage-rent.
2 R.J Rico, “Greystar to Pay $7 Million to Settle Claims over RealPage Rent‑Setting Algorithms,” AP News, November 19, 2025, https://apnews.com/article/greystar-realpage-lawsuit-settlement-rents-ded01ee0e9f1b92f38aebb82311c7470.
3 Cal. Assemb. Comm. Judiciary, “Cartwright Act: violations,” Apr. 4, 2025. https://leginfo.legislature.ca.gov/faces/billHistoryClient.xhtml?bill_id=202520260AB325
4 Cal. Assemb. Comm. Appropriations, “Cartwright Act: violations,” May 12, 2025.
5 Cal. Assemb., “Cartwright Act: violations,” Sep. 9, 2025.
6 Cal. Assemb. Judiciary, “Cartwright Act: violations,” Apr. 4, 2025.
7 Cal. Assemb. Privacy and Consumer Protection, “California Preventing Algorithmic Collusion Act of 2025,” July 14, 2025. https://leginfo.legislature.ca.gov/faces/billAnalysisClient.xhtml?bill_id=202520260SB295.
8 Cal. Assemb., “California Preventing Algorithmic Collusion Act of 2025,” Sep. 3, 2025.
9 Cal. Assemb. Comm. Privacy and Consumer Protection, “Housing rental terms: algorithmic devices,” July 14, 2025. https://leginfo.legislature.ca.gov/faces/billAnalysisClient.xhtml?bill_id=202520260SB52
10 Cal. S. Comm. Judiciary, “Housing rental terms: algorithmic devices,” Apr. 4, 2025.
11 Cal. S. Comm. Judiciary, “Preventing Algorithmic Price Fixing Act: prohibition on price-fixing algorithm use,” Apr. 8, 2025. https://leginfo.legislature.ca.gov/faces/billAnalysisClient.xhtml?bill_id=202520260SB384