New Antitrust Cases and Statements of Interests About Algorithmic Collusion


Author: Luis Blanquez

We recently wrote about the Federal Trade Commission’s blog post explaining how relying on a common algorithm to determine your pricing decisions might violate Section 1 of the Sherman Act.

The FTC has Algorithmic Price-Fixing in its Antitrust Crosshairs

It was just a matter of time until the first cases would hit the courts. That’s why during the last couple of years, we’ve seen four main federal antitrust cases alleging that algorithmic pricing might violate the antitrust laws. In three of them, the antitrust agencies also filed Statements of Interest (SOI), outlining the agencies’ opinion about what the legal principles applicable to claims of algorithmic price fixing should be.

Realpage, Inc. Software Antitrust Litigation

This multidistrict litigation in the Middle District of Tennessee involves unlawful price-fixing schemes against multifamily housing developers and managers, and student housing developers and managers, both organized by RealPage––a software algorithm company. RealPage developed software to collect property owners’ and managers’ data, used for pricing and inventory strategies, that later shared with its clients.

In January 2024, the Court: (i) denied the motion to dismiss the multifamily housing cases––the renters plausibly alleged an antitrust violation, but (ii) rejected claims alleging a horizontal price-fixing conspiracy among landlords, which would have been per se illegal. The Court, however, concluded that those same landlords vertically conspired with RealPage. The Court also dismissed the student housing plaintiffs’ complaint.

In parallel, the DOJ opened an investigation and filed a SOI. Among other things, the DOJ highlighted:

  • The fact that today software algorithms process more information more rapidly than humans and can be employed to fix prices. The technical capabilities of software can enhance competitors’ ability to optimize cartel gains, monitor real-time deviations, and minimize incentives to cheat.
  • Section 1 prohibits competitors from fixing prices by knowingly sharing their competitive information with, and then relying on pricing decisions from, a common human pricing agent who competitors know analyzes information from multiple competitors. The same prohibition applies where the common pricing agent is a common software algorithm.
  • Factual allegations in both complaints point to evidence of an invitation to act in concert followed by acceptance—evidence that is sufficient to plead concerted action. Among other things, RealPage required each user to submit real-time pricing and supply data to it, and RealPage’s marketing materials allegedly “touted” its use of “non-public data from other RealPage clients,” enabling them to “raise rents in concert”; as well as the algorithms’ ability to “facilitate collaboration among operations” and “track your competition’s rent with precision.”
  • The complaints then allege that the landlords “gave their adherence to the scheme and participated in it.” In particular, the landlords allegedly sent RealPage the non-public and competitively sensitive data (as RealPage proposed), and overwhelmingly priced their units in line with RealPage’s suggested prices (80-90%). Indeed, the complaints also contain ample allegations on how RealPage directly constrained the “deviations” from its suggested prices, including by enforcing and monitoring compliance with those prices, so the landlords effectively delegated aspects of their pricing decisions.
  • Relatedly, the multifamily plaintiffs allege that the landlords jointly delegated aspects of decision making on prices to RealPage. They allege that, by using RealPage’s pricing algorithms, each client defendant “agreed” to a common plan that involved “delegat[ing] their rental price and supply decisions to a common decision maker, RealPage.” Indeed, RealPage allegedly touted this feature—stating in a press release that it gives clients “the ability to ‘outsource daily pricing and ongoing revenue oversight,’” such that RealPage could “set prices” as though it “own[ed]” the clients’ properties “ourselves.’”
  • Jointly delegating any part of the decision-making process reflects concerted action. That the delegation is to a software algorithm, rather than a human, makes no difference to the legal analysis. Just as “surrender[ing] freedom of action. . . and agree[ing] to abide by the will of the association” can be enough for concerted action, so can be relying on a joint algorithm that generates prices based on shared competitively sensitive data.
  • The “per se” rule prohibiting price fixing applies to price fixing using algorithms. And the analysis is no different simply because a software algorithm is involved. The alleged scheme meets the legal criteria for “per se” unlawful price fixing. Although not every use of an algorithm to set price qualifies as a per se violation of Section 1 of the Sherman Act, it is per se unlawful when, as alleged here, competitors knowingly combine their sensitive, nonpublic pricing and supply information in an algorithm that they rely upon in making pricing decisions, with the knowledge and expectation that other competitors will do the same.

The District of Columbia Attorney General has also filed a similar action in the Superior Court of D.C., alleging violations of the D.C. Antitrust Act.

Duffy v. Yardi Systems, Inc.

In this case from the US District Court for the Western District of Washington, plaintiffs allege that competing landlords violated Section 1 of the Sherman Act, by unlawfully agreeing “to use Yardi’s pricing algorithms to artificially inflate” multifamily rental prices.

The Agencies also filed a SOI to explain the two legal principles applicable to claims of algorithmic price fixing. First, a competitors’ agreement to use an algorithm software with knowledge that other competitors are doing the same thing constitutes evidence of a contract, combination or conspiracy that may violate Section 1. Second, the fact that defendants deviate from the pricing algorithm’s recommendations––for instance, by just setting initial starting prices or by starting with prices lower than the ones the algorithm recommends—is not enough to get them “off the hook” for illegal price fixing (even if no information is directly shared between the parties).

The Agencies SOI’s focus was on the second point: Defendants retaining pricing discretion. The Agencies stress in the SOI that it is “per se” illegal for competing landlords to jointly delegate key aspects of their pricing to a common algorithm, even if the landlords retain some authority to deviate from the algorithm’s recommendations. Although full adherence to a price-fixing scheme may render it more effective, the effectiveness of the scheme is not a requirement for “per se” illegality. Consistent with black letter conspiracy law, the violation is the agreement, and unsuccessful price-fixing agreements are also per se illegal.

Casino-Hotel Operators Cases

Two new algorithmic pricing antitrust cases are also ongoing against casino hotel operators in Las Vegas and Atlantic City.

In Cornish-Adebiyi v. Caesar’s Entertainment, Inc., a case pending in the U.S. District Court for the District of New Jersey, plaintiffs allege a conspiracy against eight Atlantic City casino-hotel operators, and the Cendyn Group LLC, which is a provider of the algorithmic software platform, called “Rainmaker,” used to fix, raise, and stabilize the prices of casino-hotel guest rooms in Atlantic City. Rainmaker allegedly gathers real-time pricing and occupancy data to generate “optimal” room rates for each participating casino hotel, which the software then recommends to each casino hotel.

The Casino-hotel operators argued on their motion to dismiss lack of an agreement or conspiracy to increase prices. First, because there is no direct evidence of such agreement through communications or meetings among them. Second, because there is neither circumstantial evidence such as (i) all of them using Rainmaker and raising rates at the same time, nor (ii) following the algorithm’s recommendations, knowing they were sharing such information with other operators.

The DOJ and FTC also filed a SOI in this case, first, to address two legal errors from defendants in their motion to dismiss: (1) defendants’ suggestion that plaintiffs must identify direct communications between Casino-hotel operators to plausibly allege an agreement subject to Section 1 scrutiny; and (2) defendants’ argument that plaintiffs’ price-fixing claim must be dismissed because the recommendations generated by Rainmaker’s pricing algorithm are not binding.

Second, to summarize once again a non-exhaustive list of applicable legal principles required on claims of algorithmic price fixing: (i) Although direct communications among competitors can establish an agreement among them, there is no rule requiring proof of such communications; (ii) Section 1 reaches tacit as well as express agreements, and it also prohibits competitors from delegating key aspects of pricing decision making to a common entity, even if the competitors never communicate with each other directly; (iii) the Court need not apply the traditional “parallel conduct and plus factors” analysis and may infer a tacit agreement “from an invitation proposing collective action followed by a course of conduct showing acceptance” of that invitation, and (iv) an agreement among competitors to fix the starting point of pricing is per se unlawful, no matter what prices the competitors ultimately charge.

Previously, in Gibson v. MGM Resorts International, plaintiffs filed similar allegations against casino-hotel operators on the Las Vegas Strip that also allegedly used Rainmaker. In this case, however, the Agencies did not file a SOI, and the Court granted the defendants’ motion to dismiss, indicating that plaintiffs failed to show the following: (i) that defendants used the same pricing algorithm, (ii) that all casino-hotel operators used Rainmaker at or around the same time, (iii) that the casino-hotel operators exchanged nonpublic information through their use of the same software––because a successful “hub and spoke” antitrust theory based on the use of algorithmic pricing depends in part on the exchange of nonpublic information between competitors through the algorithm––a different outcome from the point made by the DOJ on the cases above, and (4) that all defendants were bound to accept the prices that the pricing software recommended to them.


Section 1 prohibits competitors from fixing prices by knowingly sharing their competitive information with, and then relying on pricing decisions from, a common software algorithm, who competitors know analyzes information from multiple competitors. It’s irrelevant that the algorithm maker isn’t a direct competitor if you and your competitors each agree to use their product knowing the others are doing the same in concert.

Jointly delegating any part of the decision-making process on an algorithm that generates prices based on shared competitively sensitive data, with the knowledge and expectation that other competitors will do the same, is enough to show concerted action. That the delegation is to a software algorithm, rather than a human, makes no difference to the legal analysis. In other words, there is no legal requirement that a plaintiff must allege specific communications directly among competitors to allege an agreement subject to Section 1.

Indeed, allegations of an agreement on parallel conduct and “plus factors” are unnecessary when a common entity such as an algorithm provider proposes a common plan such that its invitation inherently contemplates concerted action—e.g., by inviting a group of competitors to jointly delegate key aspects of pricing decision making to it. So long as the algorithm provider and its competitor clients are connected through a common agent in “a unity of purpose or a common design and understanding”, they are acting in concert.

But in light of the different outcome on the Gibson case––considering different requirements to show concerted action with a common algorithm software platform under Section 1––, it might take years to develop a solid body of black letter law.

Image by Oleg Gamulinskii from Pixabay

Contact Information