Last month, the U.S. District Court for the Northern District of California dismissed an antitrust lawsuit challenging several hotel chains’ use of AI software to suggest allegedly supra-competitive room rates. Dai v. SAS Inst. Inc., No. 24-CV-02537-JSW, 2025 WL 2078835 (N.D. Cal. July 18, 2025). The decision reinforces the specificity required to plead a conspiracy under Section 1 of the Sherman Act, including where the anticompetitive restraint is alleged to be baked into novel technologies like pricing algorithms.
Background
The pricing algorithm at issue in Dai is used in revenue management and profit optimization software (“RMS”). According to IDeaS Inc. (“IDeaS”)—the maker of the software and an additional defendant in the case—RMS “uses artificial intelligence (AI) to help hospitality professionals optimally price rooms, event spaces, and more” by “processing vast sums of data related to factors like supply and demand, local events, and historical records.” The software “automatically consider[s] all the circumstances that go into setting room prices on a given day” and then “automatically set[s] prices intended to maximize revenue.” According to the Dai plaintiffs, hotel chains enter into contracts with IDeaS to receive recommended prices, and generally implement those prices or else “use the pricing recommendations as a starting point to set prices.” See Complaint, ECF No. 1 ¶ 86, 24-CV-02537-JSW (N.D. Cal. April 26, 2024).
Allegations of Agreement
The Dai plaintiffs alleged that the defendant hotel chains conspired with IDeaS to fix hotel room prices nationwide. The complaint asserted a “hub-and-spokes” conspiracy wherein the hotel chain competitors make up the rim, with each forming a spoke through its vertical agreement with IDeaS—the hub of the conspiracy. See id. ¶¶ 96, 107, 24-CV-02537-JSW; see, e.g., In Re Musical Instruments & Equipment Antitrust Litig., 798 F.3d 1186, 1192 (9th Cir. 2015). Plaintiffs argued in opposing defendants’ motion to dismiss that the hotel chains had engaged in parallel conduct, with certain “plus factors” “support[ing] the inference of conspiracy.” ECF No. 105 at 15, 24-CV-02537-JSW (N.D. Cal. Oct. 28, 2024).
Specifically, the plaintiffs claimed that the hotel chains used IDeaS’s technology and charged supra-competitive rates based on its recommendations in parallel with each other, adopting the technology’s recommendations “in nearly every instance.” Dai, 2025 WL 2078835, at *4. As plus factors, plaintiffs claimed that the hotel chains would have been motivated to collude in this way by the economic pressures of the COVID-19 pandemic, IDeaS’s advertisements touting that the tool increases revenue by 8-15%, and the “inelastic” nature of the hotel market. Id. The plaintiffs also alleged that the hotels acted against their own economic interests by giving IDeaS confidential commercial information, knowing it would be shared with their competitors.
Analysis
The court rejected these allegations, finding them insufficient to support a hub-and-spokes theory of liability. Citing Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007), the court explained that a Section 1 complaint must include “enough factual matter (taken as true) to suggest that an agreement was made,” such that allegations “of parallel conduct and a bare assertion of conspiracy will not suffice.” Dai, 2025 WL 2078835, at *3 (quoting Twombly, 550 U.S. at 556) (cleaned up). Applying this standard, the court found the plaintiffs’ allegations of parallel conduct implausible because the complaint did not specify when any of the hotel chains began using IDeaS or when they began changing their pricing. Moreover, the facts and citations included in the complaint did not support the proposition that the hotels adopted these recommendations “in nearly every instance,” as the plaintiffs claimed. Id.
Nor was the court persuaded by the plaintiffs’ alleged plus factors. With the exception of their allegations regarding the constraints associated with the hotel market, the court found the plus factors too general and conclusory to support a reasonable inference of conspiracy. The court acknowledged that “high barriers to entry [into the hotel market] and expansion and demand for hotel rooms” “c[ould] be indicative of collusion,” but concluded that this factor was not “strong enough on its own to nudge Plaintiffs’ claims of a horizontal agreement across the line from possible to plausible.” Id. at *4.
Noting that courts analyzing similar claims have treated “the exchange of confidential information to be a decisive [plus] factor,” the court distinguished the complaint from those that survived motions to dismiss in In re RealPage, Inc., Rental Software Antitrust Litig., 709 F. Supp. 3d 478 (M.D. Tenn. 2023) (which we have previously covered here), and Duffy v. Yardi Systems, Inc., 758 F. Supp. 3d 1283 (W.D. Wash. 2024)—based on “more [extensive] detail[]” with respect to information sharing. Dai, 2025 WL 2078835, at *4. RealPage, a multidistrict litigation, initially involved two categories of plaintiffs: those alleging illegal conduct in the multifamily housing market (the “Multifamily Plaintiffs”) and those alleging illegal conduct in the student housing market (the “Student Plaintiffs”). The Multifamily Plaintiffs’ complaint survived; the Student Plaintiffs’ complaint did not. As the RealPage court put it in declining to dismiss the Multifamily Plaintiffs’ complaint, a two-way exchange of commercially sensitive information is highly persuasive indirect evidence of horizontal agreement because:
It would clearly not be in any individual Defendant’s economic self-interest to contribute its data to [the software] without knowing that it would benefit from its horizontal competitors doing the same. Put another way, the contribution of sensitive pricing and supply data for use by [the software] to recommend prices for competitor units is in Defendants’ economic self-interest if and only if Defendants know they are receiving in return the benefit of their competitors’ data in pricing their own units.
709 F. Supp. 3d at 512.
Indeed, the surviving complaints in RealPage and Duffy specifically pleaded that the subject software worked both by inputting nonpublic competitor data into the software and by generating recommendations based on that same data—thereby ensuring that the competitor-defendants benefitted from the otherwise economically irrational act of sharing their private commercial information with competitors. See Duffy, First Amended Complaint, ECF No. 113 at 71, 23-cv-1391-RSL (W.D. Wash. Nov. 3, 2023) (alleging that users of the subject software “appreciated the fact that if they shared data, they would get data from other clients . . . so that everybody was benefitting from the data”) (emphasis added); RealPage, Second Amended Complaint, ECF No. 530 ¶ 13, 23-md-03071 (M.D. Tenn. Sept. 7, 2023) (alleging that the defendants “provide[d]” software “with vast amounts of their non-public proprietary data” which “[wa]s fed into a common data pool” used to train the software’s algorithm and make pricing recommendations).
The plaintiffs in Dai tried to analogize those cases to theirs, quoting a 2017 statement by former Acting Chair of the Federal Trade Commission Maureen Ohlhausen claiming that antitrust law “prohibits using an intermediary”—whether in the form of a third party, or an algorithm—“to facilitate the exchange of confidential business information.” However, the court found that concept inapposite, as the complaint only alleged that the subject software “plugs” such information “into its algorithm,” without specifying whether or how this information is incorporated into the pricing recommendations used by competitors. Dai, 2025 WL 2078835, at *5. Absent such allegations, the Dai court found the plaintiffs’ more general allusions to the software’s use of confidential information insufficient to support a Section 1 claim.
Other Dismissals of Alleged AI Price Fixing
Dai is consistent with a growing body of precedent largely rejecting allegations of collusion and anticompetitive harm in the context of AI pricing software in the housing and hotel industries. For example:
- Cornish-Adebiyi v. Caesars Entm't, Inc. granted a motion to dismiss Section 1 conspiracy allegations against hotel-casinos on the grounds that plaintiffs failed to plead “how” their “mere use of the specific software . . . [was] suggestive of culpable conspiracy” and that they did not explain how non-public room price and occupancy data “is used” by the software “once it is handed over.” No. 1:23-CV-02536-KMW-EAP, 2024 WL 4356188 (D.N.J. Sept. 30, 2024).
- Gibson v. MGM Resorts Int'l (“Gibson I”) dismissed (with leave to amend) an alleged Section 1 conspiracy against a different set of hotel chains, because the plaintiffs failed to “explicitly” allege whether, through the subject software, the defendants “ever receive[] confidential information belonging to another [defendant]” and “whether the pricing recommendations ‘generated’ to [defendants] include that confidential information fed in.” The court noted that the complaint could be read to suggest that the defendants “only get their own confidential information back, mixed with public information from other sources.” According to the court, this “ambiguity” rendered the plaintiffs’ hub-and-spoke theory of liability insufficient. No. 2:23-cv-00140-MMD-DJA, 2023 WL 7025996 (D. Nev. Oct. 24, 2023).
- Gibson v. Cendyn Grp., LLC (“Gibson II”) dismissed the Gibson I plaintiffs’ amended complaint with prejudice on similar grounds. The court reiterated: “Plaintiffs’ failure to plausibly allege the exchange of confidential information from one of the spokes to the other through the hub’s algorithms is a[] fatal defect with their [Section 1] claim because it . . . compels the conclusion that there is no rim.” No. 2:23-cv-00140-MMD-2JA, 2024 WL 2060260 (D. Nev. May 8, 2024).
As previously mentioned, In re RealPage, Inc., Rental Software Antitrust Litig., 709 F. Supp. 3d 478 (M.D. Tenn. 2023), dismissed the Student Plaintiffs’ complaint even though the Multifamily Plaintiffs’ complaint survived. The court rejected the Student Plaintiffs’ complaint not because of any deficiencies related to the alleged parallel conduct or plus factors, but because it failed to plausibly allege market power over 27 alleged regional submarkets throughout the United States. Id. at 530. While the Multifamily Plaintiffs’ case has continued to move forward, on August 8, 2025, one of the defendants, Greystar Real Estate Partners, LLC, announced that it reached a settlement in principle with the Multifamily Plaintiffs, as well as with the Department of Justice with respect to a lawsuit alleging similar Section 1 violations that is currently pending in the Middle District of North Carolina.
Conclusion
Given that the above cases all concerned similar software and overlapping alleged markets, yet saw different outcomes, the Dai dismissal highlights the level of detail required to successfully plead a “hub-and-spokes” conspiracy in the algorithmic-pricing context. Plaintiffs may need to include detailed allegations about how the technology at issue works in order to plead parallel conduct and plus factors that take it across the line to plausibility.
