The Algorithmic Price-Fixing Revival: Why the Third Circuit’s Atlantic City Casino Ruling Opens a New Front in Antitrust Class Actions

The Third Circuit’s decision reviving a proposed antitrust class action against Atlantic City casino operators is the most consequential civil litigation development in this week’s appellate docket — not because of what it says about hotel room pricing, but because of what it signals about the judiciary’s willingness to entertain algorithmic collusion theories at the pleading stage. The court held that plaintiffs adequately alleged a violation of Section 1 of the Sherman Act, 15 U.S.C. § 1, based on the casinos’ shared use of AI-powered revenue management software that allegedly facilitated coordinated pricing without traditional evidence of an agreement. That holding, if it survives further review, could reshape antitrust class action practice across every industry where pricing algorithms have become standard infrastructure.

The Holding and Its Immediate Context

The case, which had been dismissed below, alleges that Atlantic City casino-hotel operators used a common AI-driven revenue management platform to set room rates, effectively allowing the software to function as a hub through which competitively sensitive pricing information flowed and from which supracompetitive prices emerged. The district court found the allegations insufficient to state a plausible claim under the standard established by Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007), reasoning that the mere parallel use of a common software tool did not adequately allege an agreement — the sine qua non of Section 1 liability.

The Third Circuit disagreed, concluding that the plaintiffs’ allegations — which reportedly included details about how the software ingested competitors’ non-public pricing data and generated pricing recommendations that the casinos adopted at unusually high rates — crossed the plausibility threshold. The court did not hold that algorithmic pricing is per se illegal, nor did it resolve whether the defendants actually reached an agreement. It held only that the complaint alleged enough to proceed to discovery. But in antitrust class action practice, surviving a motion to dismiss is often the ballgame. Discovery costs and the in terrorem effect of class certification exposure create enormous settlement pressure, and defendants know it.

Why This Matters Beyond Atlantic City

The decision arrives at a moment when algorithmic pricing litigation is proliferating across sectors. The Justice Department’s Antitrust Division has been aggressively pursuing the theory that competitors who delegate pricing decisions to a common algorithm may violate Section 1 even absent direct communications — a position articulated most prominently in the government’s statement of interest in the RealPage multifamily housing litigation, where property managers allegedly used RealPage’s revenue management software to coordinate rental prices. The Third Circuit’s reasoning tracks that enforcement posture and gives it judicial imprimatur at the appellate level.

For defense counsel, the ruling complicates the standard playbook. The conventional motion-to-dismiss strategy in algorithmic pricing cases has been to argue that parallel adoption of a common tool is merely conscious parallelism — lawful under Theatre Enterprises, Inc. v. Paramount Film Distributing Corp., 346 U.S. 537 (1954) — and that plaintiffs must allege “plus factors” tending to exclude independent action. The Third Circuit’s willingness to treat the nature of the software itself — specifically, its capacity to aggregate and act upon competitors’ confidential data — as a plus factor significantly lowers the pleading bar. If other circuits adopt similar reasoning, the Twombly/Iqbal framework that has served as the defense bar’s primary gatekeeping tool for nearly two decades may prove far less effective in this emerging category of cases.

The Class Certification Horizon

The harder fight, of course, will come at class certification under Federal Rule of Civil Procedure 23(b)(3), where plaintiffs must demonstrate that common questions predominate. The Ninth Circuit’s recent decision blocking class treatment in the Tesla self-driving litigation — on the ground that individualized questions about what each buyer saw before purchase overwhelmed any common issues — illustrates how certification remains the true bottleneck in consumer class actions. But antitrust cases involving algorithmic pricing may present a different calculus. If the alleged mechanism of harm is a common software platform applied uniformly across a market, the predominance inquiry may actually favor plaintiffs more than in typical consumer fraud cases, where individual reliance and exposure questions tend to fragment the class.

That dynamic makes the Third Circuit’s ruling doubly significant: it not only revives the case but sets up a certification fight on terrain that may be more favorable to plaintiffs than the consumer protection context. Defense counsel in the RealPage MDL and similar pending matters should be reassessing their certification strategies now.

The Broader Appellate Landscape

The Third Circuit’s decision is part of a strikingly active period for class action litigation across the federal circuits. The Tenth Circuit’s divided ruling that British American Tobacco must face class claims over Natural American Spirit’s “additive-free” marketing — with a dissent explicitly calling for Supreme Court review — suggests that the boundaries of Rule 23 remain contested terrain. The Eighth Circuit’s dismissal of the appeal challenging objections to Bayer’s $7.25 billion Roundup settlement leaves intact a state-court resolution mechanism that mass tort practitioners have been watching closely. And the Supreme Court’s refusal to hear Zillow’s bid to escape its investor class action preserves the Ninth Circuit’s pro-certification result in the securities context.

Taken together, these decisions paint a picture of an appellate judiciary that is actively recalibrating class action doctrine across multiple substantive areas simultaneously — antitrust, consumer protection, securities, mass torts, and privacy. The era of assuming that Twombly and Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338 (2011), would inexorably contract class action practice appears to be giving way to something more nuanced: a doctrine that is tightening in some contexts while opening in others, particularly where novel technologies create new mechanisms of alleged collective harm.

Practice Implications

For plaintiffs’ counsel, the message is clear: algorithmic collusion complaints that include specific factual allegations about how the software works — what data it ingests, how it generates recommendations, and how uniformly those recommendations are adopted — can survive dismissal. Vague assertions that competitors “used the same software” will not suffice, but detailed technical allegations about the software’s data architecture may function as the plus factors that Twombly demands.

For defense counsel, the ruling underscores the urgency of front-loading factual development about the software’s actual operation. If the algorithm does not, in fact, aggregate competitors’ confidential data or if adoption rates of its recommendations vary significantly, those facts need to be established early — ideally at the pleading stage through judicial notice or incorporation by reference, or failing that, through targeted early discovery motions.

Both sides should monitor whether the defendants seek en banc or Supreme Court review. Given the circuit split potential — particularly if the more defense-oriented circuits weigh in on similar algorithmic pricing theories — this case has the architecture of a certiorari candidate. The intersection of antitrust law and artificial intelligence is precisely the kind of novel legal question that the current Court has shown appetite for addressing.