The Third Circuit’s decision in Cornish-Adebiyi v. Caesars Entertainment, Inc. underscores how alleged data pooling can heighten antitrust risk and yield treble-damages exposure.1 Companies considering adopting pricing software tools should be mindful of these risks.
The Third and Ninth Circuits split on parallel AI pricing implementations
Plaintiffs assert algorithmic pricing collusion claims as “hub-and-spoke” conspiracies. In a hub-and-spoke model, a central entity (the “hub”) facilitates anticompetitive coordination among competing businesses (the “spokes”). In developing cases, plaintiffs allege that competitors use shared pricing software as an intermediary to exchange—and, importantly, align—their prices.
Courts tend to allow these claims to reach discovery when plaintiffs plausibly allege facts suggesting that rival users of the software agreed with one another to set prices through the software. The theory is that such an alleged agreement among competitors forms the requisite “rim” of the hub-and-spoke conspiracy.
One of the earliest cases in the current wave of AI pricing litigation is Gibson v. Cendyn Group, a case involving alleged collusion among Las Vegas hotels using pricing software developed by Cendyn Group.2 The district court dismissed the complaint for failing to plausibly allege an overarching collusive agreement among the hotel defendants. Although the hotels each entered licensing arrangements with the same software vendor, the complaint lacked circumstantial facts indicating that the hotels agreed among themselves to set rates.
The District of New Jersey dismissed a similar case alleging collusion among Atlantic City hotels using the same software in Cornish-Adebiyi.3 With Gibson as a guide, the writing seemed to be on the wall for the Cornish-Adebiyi plaintiffs. But on July 29, 2026, the Third Circuit reversed.
Why did the Third Circuit reach the opposite conclusion from the Ninth Circuit when evaluating allegations about the use of the same software in the same industry? The distinction boils down to the plausibility of the plaintiffs’ allegations that the pricing software commingled rivals’ private data to generate individual pricing recommendations. Companies face greater antitrust risk when using a pricing tool that not only collects proprietary information from its users but also pools that information to calculate recommended prices.
The Third Circuit saw data pooling where the Ninth Circuit did not
The District of Nevada in Gibson found fatal defects in the alleged rim. Critically, the court found no factual allegation that the hotels exchanged confidential data through the software. The hotels sent data to the vendor, but the plaintiffs did not allege that the vendor commingled the hotels’ data when calculating rate recommendations.
Without data pooling, the court held that the plaintiffs had not alleged that any hotel received a rate recommendation based on its rivals’ confidential data. The complaint therefore failed to allege an overarching rate agreement among the hotels. The plaintiffs did not contest the absence of data pooling on appeal, and the Ninth Circuit affirmed, holding that merely using the same software was an insufficient basis to infer an agreement.
The District of New Jersey dismissed the Cornish-Adebiyi complaint on the same grounds. Each hotel provided non-public room pricing and occupancy data to the same provider, but the court found no allegation that the data was thereafter pooled to generate pricing recommendations. The court therefore found no alleged exchange of confidential information and, consequently, no sufficient basis to infer that the hotels agreed among themselves to align rates through the software.
The Third Circuit disagreed. The plaintiffs’ complaint alleged that the algorithm functions by continuously analyzing hotels’ non-public real-time information “along with the same type” of information submitted by their competitors. The court pointed to the complaint’s allegations that each hotel had a common understanding that its recommended rates relied on competitors’ non-public data.
That awareness, the court held, supported an inference of the hotels’ agreement to align on prices based on recommendations from software acting as “the hub for their collective data.” Rebuffing what it characterized as an excessively demanding pleading standard imposed below, the court explained that plaintiffs need not detail precisely how the algorithm facilitated information sharing to state a claim.
The divergent outcomes thus appear to have turned not on a disagreement over whether information pooling matters, but on the sufficiency of the pooling allegations. The Cornish-Adebiyi plaintiffs alleged a mutual understanding of non-public data exchange; the Gibson plaintiffs did not.
Staggered software adoption
The Third and Ninth Circuits also disagreed on the significance of the users’ staggered adoption of the pricing software. Gibson held that the defendants’ deployment of the software at different times over approximately ten years reinforced the implausibility of any alleged anticompetitive agreement among them.
The Third Circuit reached a different conclusion in Cornish-Adebiyi. The defendants adopted the software at various points over a 14-year span, but the court held that this did not defeat an inference of collusion.
As the court saw it, the alleged conspiracy did not begin when each defendant first licensed the software. The relevant alleged conduct was instead the defendants’ continuous use of the same software coupled with synchronous pricing movements. The Third Circuit held that parallel conduct—with the alleged information pooling—was sufficient to infer an agreement at the pleading stage.
Retaining pricing authority
Customer discretion to accept or reject pricing recommendations is also an important factor in algorithmic pricing litigation. Gibson highlighted that the hotels often ignored Cendyn’s recommendations. In Cornish-Adebiyi, by contrast, the Third Circuit held that an alleged 90% adherence rate among the hotels supported an inference of agreement at the pleading stage.
Takeaway
These cases identify important considerations for businesses that employ algorithmic pricing tools. Using the same tool as a rival is not by itself an antitrust concern, even if each company contributes competitively sensitive data. But how the software vendor uses that data to inform prices is crucial to evaluating risk.
Courts are more likely to infer an anticompetitive agreement among rivals when a plaintiff alleges that each rival knew its confidential data was a factor in its competitors’ recommended prices. This alleged reciprocal use of confidential information is increasingly what separates complaints that survive motions to dismiss from those that do not.
A&O Shearman’s antitrust team regularly advises clients on recent developments in algorithmic pricing and has deep expertise in counseling companies on antitrust risk and compliance.
Footnotes
1 Cornish-Adebiyi v. Caesars Entertainment, Inc., No. 24-3006 (3d Cir. July 29, 2026).
2 Gibson v. Cendyn Group, LLC, No. 2:23-cv-00140 (D. Nev. May 8, 2024).
3 Cornish-Adebiyi v. Caesars Entertainment, Inc., No. 1:23-cv-02536 (D.N.J. Sept. 30, 2024).
