The recent trend of the antitrust actions targeting pricing algorithms—mainly concentrating in the hospitality and real estate industries—has reached the appellate level in two cases this year, both involving the revenue management products sold by Rainmaker/Cendyn to hotel owners.
Pending before the Ninth Circuit is the Las Vegas Strip hotel case discussed in our previous article, Gibson v. Cendyn Group, LLC, No. 24-3576 (9th Cir.). The district court in Nevada dismissed the amended complaint on both counts with prejudice. The court first reasoned that the complaint failed to plausibly allege a hub-and-spoke conspiracy to fix price, relying heavily on the lack of allegations that the defendants pooled confidential data into the revenue management software to inform price recommendations. The court then held that the complaint failed to plausibly allege the vertical agreements between Rainmaker/Cendyn and the hotel defendants constituted a “restraint of trade” under Section 1 of the Sherman Act, pointing to the lack of allegations that the hotel defendants had agreed to restrain their ability to independently price the casino hotel rooms via entering the licensing agreements. See Gibson v. Cendyn Grp., LLC, No. 2:23-CV-00140-MMD-DJA, 2024 WL 2060260 (D. Nev. May 8, 2024). The plaintiffs appealed the decision.
The case is now fully briefed and pending oral argument before the Ninth Circuit, with the Department of Justice (DOJ) submitting an amicus brief in support of the plaintiffs-appellants. Interestingly, the plaintiffs initially appealed the district court’s decision on both counts, but in their reply brief abandoned the appeal on the hub-and-spoke count, apparently recognizing the relative weakness of their per se violation claim based on the alleged facts in the complaint. See Reply Br. at 32, Gibson v. Cendyn Group, LLC, No. 24-3576 (9th Cir. Feb. 7, 2025). As a result, the only count that is waiting for the appellate court’s decision is the alleged rule-of-reason violation through the vertical agreements and the key issues there are whether vertical agreements by themselves constitute a “restraint of trade” and whether alleging a rise in price is sufficient to satisfy the anticompetitive effect requirement for constituting an “unreasonable” restraint of trade under the rule of reason framework.
The defendants-appellees in their answering brief distinguished their case from the parallel antitrust cases involving the rental property revenue management software sold by RealPage and Yardi to commercial landlords. They argued that in those cases, most of which survived motions to dismiss, the software in question allegedly used non-public, proprietary data fed into the software by the landlord customers to generate recommendations for rents and to share that information among competitors. Answering Br. at 33-34, Gibson v. Cendyn Group, LLC, No. 24-3576 (9th Cir. Feb. 7, 2025). We wrote about those cases in our prior article, and we anticipate new lawsuits and/or enforcement actions will be filed at the district court level that emulate the allegations in the complaints of the cases surviving motions to dismiss. For example, the DOJ filed an enforcement action against RealPage last year and recently amended its complaint to include six landlords as co-defendants, making similar allegations as the private lawsuits against RealPage. See Am. Compl., United States v. RealPage Inc., No. 1:24-cv-00710 (M.D.N.C. Jan. 7, 2025).
Another dismissal ruling waiting for appellate review is Cornish-Adebiyi v. Caesars Entertainment, Inc., No. 24-03006 (3d Cir.), pending before the Third Circuit. There, a similar group of plaintiffs-appellants appealed the New Jersey district court’s dismissal of their complaint against casino hotel owners in Atlantic City alleging a price-fixing conspiracy through the hotel owners’ alleged common use of the same revenue management products at issue in Gibson. The New Jersey district court cited the dismissal opinion in Gibson at length to support its decision. See Cornish-Adebiyi v. Caesars Ent., Inc., No. 1:23-CV-02536-KMW-EAP, 2024 WL 4356188 (D.N.J. Sept. 30, 2024).
The plaintiffs-appellants in Caesars submitted their opening brief in January, and distinguished their case from Gibson in that their complaint did contain allegations that the software at issue used confidential and proprietary information to train artificial intelligence and machine learning, which was then used to generate pricing recommendations shared among the competing casino hotels. Opening Br. at 21-22, Cornish-Adebiyi v. Caesars Entertainment, Inc., No. 24-03006 (3d Cir. Jan. 21, 2025). They also argued that plaintiffs are not “required at the pleading stage to show how the Rainmaker algorithms function. The collective delegation of pricing authority to Rainmaker itself is sufficient.” Id. at 32 (emphasis in the original).
A key issue common to all these cases is how the revenue management software at issue “crunch” the confidential data it receives from its users. As a matter of law, can it be said that competitors are sharing their confidential information if none of them can “back out” another competitor’s confidential data from the algorithm that generates pricing recommendations, even if the algorithm is “trained” by using the confidential data? Or is an algorithm automatically “tainted” unless every source or input of data is public and non-proprietary?
The plaintiffs-appellants in both Gibson and Caesars quoted an analogy made by a former Federal Trade Commission (FTC) chair, Maureen K. Ohlhausen, in 2017: “Everywhere the word ‘algorithm’ appears, please just insert the words ‘a guy named Bob.’ Is it okay for a guy named Bob to collect confidential price strategy information from all the participants in a market, and then tell everybody how they should price? If it isn’t ok for a guy named Bob to do it, then it probably isn’t ok for an algorithm to do it either.” Opening Br. at 3, Gibson v. Cendyn Group, LLC, No. 24-3576 (9th Cir. Feb. 7, 2025); Opening Br. at 10, Cornish-Adebiyi v. Caesars Entertainment, Inc., No. 24-03006 (3d Cir. Jan. 21, 2025).
The defendants-appellees in Gibson countered with the following analogy: “Indeed, alleging that the algorithms improved through machine learning is no different than an attorney improving her skills over time with the benefit of . . . confidential client information she gains with each client engagement, without ever sharing any one client’s confidential information with any other client.” Answering Br. at 36, Gibson v. Cendyn Group, LLC, No. 24-3576 (9th Cir. Feb. 7, 2025) (internal citation and quotation marks omitted).
The two appellate cases will play a key role in determining which of these two visions of the role of artificial intelligence or machine learning plays in pricing will prevail. Regardless of the results, we anticipate antitrust litigation around pricing algorithms will continue to be active in 2025.
Defendants in these high-stakes, years-long antitrust lawsuits are often simultaneously fighting a separate battle with their insurers to cover the defense costs and seek indemnification for any potential liabilities. In our prior article, we discussed two cases where the landlords in the RealPage cases filed lawsuits against their insurers over denial of coverage, and cautioned policyholders to carefully review the exclusionary language in their commercial policies to best position themselves should they face antitrust lawsuits.
At the end of last year, the insurer in one of those two cases was granted summary judgment, resulting in the dismissal of the insurance coverage lawsuit. Willow Bridge Prop. Co., LLC v. Arch Specialty Ins. Co., No. 3:24-CV-0029-D, 2024 WL 5240154, at *9 (N.D. Tex. Dec. 27, 2024). The court’s decision did not focus on the otherwise broad and ambiguous exclusion in the professional liability policy that would bar coverage over “any Claim for, based upon, arising from, or in any way related to unfair trade practices or violation of consumer protection laws,” which arguably does not squarely exclude antitrust actions. Instead, the court held that coverage was not triggered in the first place, relying on a narrow interpretation of the defined terms of “Wrongful Act” and “Personal Injury” in the policy. Id. at *6-8.
In another insurance coverage lawsuit brought by a defendant in the broiler chicken consolidated antitrust class action (another “hub-and-spoke” antitrust litigation, albeit not involving allegations of artificial intelligence or pricing algorithm), the California district court granted summary judgment in favor of the excess insurer, rejecting the policyholder’s argument that the antitrust exclusion in the liability insurance policy did not bar coverage because many of the claims in the underlying litigation asserted unjust enrichment, violations of state consumer protection laws, fraud, and false advertising, which were apparently not antitrust claims. Beazley Ins. Co., Inc. v. Foster Poultry Farms, No. 1:21-CV-01806-LHR-SKO, 2024 WL 3904994, at *5 (E.D. Cal. Aug. 22, 2024). The court reasoned that while California law imposed a duty to defend the entire case as long as the underlying litigation presented a potential of covered liability, the non-antitrust claims described above were all alleging the same facts as the antitrust claims and their success was predicated on the success of the antitrust claims. See id. at *6-8.
In contrast, when the insurance policy explicitly includes antitrust coverage, courts seem to be more willing to protect that coverage, even if other provisions in the policy apparently take that coverage away. In a recent decision by the Tenth Circuit, the appellate court held that when the policy at issue expressly covered “claims for price fixing and monopolization,” an exclusion in the same policy that excluded claims arising from any activities involving the core business of the policyholder was ambiguous and should be interpreted in favor of coverage, because “any scenario where price fixing or monopolization could exist” necessarily involved the policyholder’s core business activities. Allied World Specialty Ins. Co. v. Blue Cross & Blue Shield of Kansas, Inc., No. 23-3130, 2024 WL 4969268, at *4 (10th Cir. Dec. 4, 2024). In other words, a policy cannot explicitly grant coverage over antitrust claims and take away that coverage through an exclusion except through “clear, unmistakable language.” Id.
* * *
For companies and industries that rely heavily on third-party revenue management tools and pricing algorithms for their business, it is advisable to review with their insurance brokers and coverage counsel the existing policies to determine whether and how much coverage is available against any potential antitrust actions, and consider securing adequate coverage at policy renewals. They should also be on the lookout for major antitrust decisions that are anticipated this year, as that might impact how they use machine learning and artificial intelligence to conduct business going forward.
