Third Circuit Rules Shared Pricing Algorithms Can Constitute Illegal Price-Fixing, Threatening AI Business Models

With a 90% compliance rate to an algorithm's pricing recommendations, the U.S. Court of Appeals for the Third Circuit has delivered a landmark antitrust ruling that fundamentally threatens the business models of artificial intelligence pricing software. On July 29, 2026, the appellate court unanimously reversed a lower court's dismissal of a class-action lawsuit against several Atlantic City casino-hotels and their shared revenue-management vendor, Cendyn Group.[2][7]

The plaintiffs allege that the casinos used Cendyn's "Rainmaker" algorithm to orchestrate a massive price-fixing conspiracy, artificially inflating hotel room rates by 22% to 25% during a period when occupancy actually declined by 5% to 8%. By allowing the case to proceed to discovery, the Third Circuit has signaled that the deployment of shared algorithmic pricing tools can constitute an illegal horizontal agreement under Section 1 of the Sherman Act.[3][5]

For business leaders and software developers, the practical stakes are immediate and severe. If a company uses third-party AI software that pools confidential competitor data to recommend prices, it could now face class-action antitrust liability in jurisdictions following this ruling, fundamentally challenging how modern dynamic pricing tools operate across the economy.[1][2]

The mechanics of the alleged conspiracy center on the pooling of highly sensitive corporate data. According to the complaint, each participating casino-hotel continuously uploaded real-time, non-public pricing and occupancy information into the Rainmaker platform.[4][8]

Rainmaker then aggregated this confidential information with identical data streams from competing casinos. Using this collective pool of proprietary inputs, the artificial intelligence generated "optimal" room prices multiple times per day, which were automatically fed back into each hotel's reservation system.[6]

In antitrust law, this structure is known as a "hub-and-spoke" conspiracy. The software vendor acts as the central hub, while the competing businesses serve as the spokes. The critical legal question is whether there is a "rim" connecting the spokes—a horizontal agreement among the competitors to fix prices, rather than just independent, vertical agreements with the software provider.[2][3]

The U.S. District Court for the District of New Jersey initially dismissed the case in September 2024, concluding that the conspiracy lacked this essential rim. The lower court reasoned that each casino independently subscribed to Rainmaker at different times, retained final authority over their own rates, and were theoretically free to reject the algorithm's recommendations.[5]

The Third Circuit vehemently disagreed with this assessment, finding that the plaintiffs had plausibly alleged both parallel conduct and the necessary "plus factors" to infer a conspiracy. The appellate panel emphasized that the casinos were not merely using the same software, but were actively exchanging commercially sensitive information through a shared de facto pricing agent.[4][6]

A crucial factor in the Third Circuit's reversal was the staggering compliance rate. The plaintiffs alleged that the casino-hotels accepted Rainmaker's pricing recommendations approximately 90 percent of the time.[3][5]

Furthermore, the court noted that Cendyn allegedly engineered the software to discourage deviations from its suggested rates. Overriding the algorithm required special permissions, and the system actively tracked and scored hotels based on how frequently they departed from the AI's optimal prices, creating a structural enforcement mechanism for the cartel.[4][8]

The ruling effectively endorses a popular analogy among antitrust scholars: if a human being named "Bob" could not legally collect confidential pricing data from competitors and dictate what they should charge, replacing Bob with a sophisticated algorithm does not immunize the conduct from antitrust scrutiny.[8]

This decision creates a stark circuit split within the federal judiciary. Just last August, the Ninth Circuit reached the exact opposite conclusion in Gibson v. Cendyn Group, affirming the dismissal of a nearly identical lawsuit involving Las Vegas hotels and the same Rainmaker software.[2][5]

The Ninth Circuit ruled that independent decisions by competing hotels to license the same pricing software, even if it resulted in higher prices, were insufficient to state a price-fixing claim. The Supreme Court declined to review the Gibson case in April 2026, leaving the Ninth Circuit's defendant-friendly standard intact out West.[2]

Legal analysts suggest the divergent outcomes may stem from a specific factual distinction in the pleadings. The Cornish-Adebiyi plaintiffs explicitly alleged that the software pooled non-public competitor data to generate its recommendations, whereas the Gibson plaintiffs reportedly failed to emphasize that critical data-mingling mechanism.[5]

Regardless of the nuances, the Third Circuit's permissive pleading standard makes it substantially easier for plaintiffs to survive early dismissal motions in jurisdictions that follow its reasoning. This exposes companies across the hospitality, real estate, and retail sectors to massive class-action discovery costs.[2][3]

The ruling aligns closely with the aggressive enforcement posture of the Department of Justice and the Federal Trade Commission. Both agencies have recently filed statements of interest in similar algorithmic pricing cases, arguing that the joint use of common pricing algorithms and information exchange violates the Sherman Act.[2]

For artificial intelligence developers, the decision draws a perilous red line. While the court acknowledged that algorithmic pricing is not inherently anticompetitive, systems that ingest proprietary competitor data to output shared pricing strategies are now legally toxic in the Third Circuit.[3][4]

Software vendors may need to fundamentally re-architect their models, ensuring that algorithms are trained solely on public data or strictly siloed proprietary data, rather than cross-pollinating confidential inputs across competing clients. The Cornish-Adebiyi case will now plunge into the costly and invasive factual discovery phase, while the glaring circuit split makes this issue a prime candidate for eventual Supreme Court intervention.[2][6][7]