by Malcolm Girand*
Section 1 of the Sherman Act reaches only concerted action, whereas parallel outcomes that reflect independent business judgment remain lawful. The proliferation of shared algorithmic pricing tools has strained that line, and courts have divided over what a plaintiff must plead to establish an unlawful agreement. Some have dismissed these claims at the threshold, treating the absence of a clearly alleged agreement as dispositive and requiring allegations that virtually negated independent explanations at the pleading stage. This Contribution argues that such an approach misreads Bell Atlantic Corp. v. Twombly, importing an evidentiary demand associated with trial and summary judgment into Rule 12. It further argues that courts should recognize reciprocal algorithmic data-pooling—competitors contributing their own nonpublic, competitively sensitive, current data to a common pricing system that uses those data in generating recommendations for known rivals, with each contributor aware of the reciprocal arrangement—as a plus factor to infer agreement from parallel conduct. Because such reciprocal algorithmic data-pooling is difficult to explain as ordinary unilateral conduct and strongly suggests mutual assurance, it ordinarily suffices—joined with parallel conduct and the remaining elements of the claim—to support a plausible inference of agreement and reach discovery.
Suppose a group of competitors each quietly handed their pricing data to the same consultant, who studied it and told each firm what to charge. Each firm knew that its rivals were supplying the same data and taking the same advice: the entire arrangement contemplated reciprocal participation. That structure supports an inference of horizontal agreement even though the competitors never met, never spoke, and dealt only with the consultant. The Supreme Court settled as much in Interstate Circuit, when it inferred agreement among distributors who accepted a common plan knowing the others were in and knowing the plan worked only if they were in too.1 The Sherman Act—the foundational federal antitrust statute—polices the substance, not the form of coordination.2
Replace the consultant with an algorithm performing a similar function, and courts hesitate. Section 1 of the Sherman Act prohibits contracts, combinations, and conspiracies that unreasonably restrain trade.3 But Section 1 condemns only conduct that “stem[s] from” an agreement, “tacit or express”;4 whereas conscious parallelism—independent firms recognizing their interdependence and moving together—remains lawful, however lockstep the result.5 Shared pricing software sits uncomfortably between those poles, and the early returns diverge sharply. The Ninth Circuit, in the first appellate decision on the question, affirmed dismissal of claims against Las Vegas hotels that licensed a common revenue-management tool.6 District courts confronting related arrangements in rental housing and health insurance markets sustained the claims and sent the parties to discovery.7 Most recently, the Third Circuit has deepened this divide. In Cornish-Adebiyi, it reversed the dismissal of claims against casino-hotels using the same software at issue in Gibson, holding that allegations of pooled non-public data, known mutual participation, and near-uniform adherence plausibly plead a horizontal agreement.8 The question is not whether pricing algorithms are unlawful—they are not—but what a plaintiff must plead to test, through discovery, whether competitors have used one to fix prices.
This Contribution argues that the dismissing courts have misapplied Twombly, the Supreme Court decision requiring a complaint to plead a “plausible” claim before the parties may proceed to discovery.9 By effectively requiring allegations approaching an express adherence agreement, or allegations that come close to negating independent explanations at the threshold, they have imported into Rule 12—the Federal Rule of Civil Procedure governing motions to dismiss—an evidentiary burden the Supreme Court has associated with proof at trial and summary judgment. This Contribution further argues that the correction is not merely replacing “plausible” with a softer adjective. Courts should recognize reciprocal data-pooling—competitors contributing their own nonpublic, competitively sensitive, current data to a common pricing system that uses those data in generating recommendations for known rivals, with each contributor aware of the reciprocal arrangement—as a plus factor. Joined with parallel conduct and the remaining elements of the claim, reciprocal data-pooling ordinarily suffices to support a plausible inference of agreement. Treating reciprocal data-pooling as a plus factor rests on one of the most probative inferences in the plus factor canon, it draws a line that recent practice suggests is administrable, and it accomplishes judicially what state legislatures are now attempting, unevenly, by statute.
Begin with the framework the algorithm cases inherit. Because conspirators rarely reduce their understandings to writing, an agreement may be inferred from circumstantial evidence, but only from parallel conduct joined to “plus factors” that make conspiracy a plausible explanation rather than a merely possible one.10 The plausibility demanded is Twombly’s, and Iqbal is explicit that plausibility “is not akin to a probability requirement.”11 A materially more demanding evidentiary standard—evidence that “tends to exclude the possibility” of independent action—derives from Monsanto and Matsushita and is associated with proof at trial or summary judgment, after discovery, not with the pleadings.12 Among the recognized plus factors, courts often describe one as especially probative: conduct that would be unusually risky, or difficult to explain, unless the actor’s rivals were acting in concert.13 Keep that hierarchy in view; the algorithm cases turn on it.
Three questions travel together in these cases and should be kept apart. The first is agreement—whether the complaint plausibly alleges concerted action at all—and it is the only question the proposed factor addresses. The second is characterization: whether an adequately alleged agreement draws per se condemnation or rule-of-reason scrutiny, a question on which the sustaining courts have themselves divided and need not be resolved here.14 The third is the possibility that an information exchange may independently violate Section 1 under the rule of reason even where no price-fixing agreement is adequately pleaded—a distinct theory with its own doctrine.15 Nothing that follows treats every exchange of sensitive data as itself a per se price-fixing agreement. The proposed plus factor, rather, speaks only to the inference of agreement.
The courts that dismissed reciprocal algorithmic data-pooling claims honored the Section 1 framework in name but inverted it in application. In Gibson, the Ninth Circuit acknowledged that it would “undoubtedly violate Section 1” for competing hotels to agree among themselves to abide by a third party’s pricing recommendations.16 The court then affirmed dismissal, treating the absence of a clearly alleged adherence pact as effectively dispositive and observing that antitrust law “does not require a business to turn a blind eye” to information its competitors also possess.17 The court reserved the harder case, noting that the analysis “might change” had the vendor shared each hotel’s confidential information among the licensees.18 Other courts extended the dismissals to complaints that gestured at data-sharing without pleading its mechanics, including the Cornish-Adebiyi district court.19 There is a defensible core here: the district court in Gibson could not tell whether hotels received anything beyond their own data back, “mixed with public information from other sources.”20 The Third Circuit has since repudiated that demand, holding that requiring plaintiffs to plead how the algorithm functions is “tantamount to expecting Plaintiffs to explain how [Defendant’s] proprietary software works without affording the discovery required to do that.”21
But the method these decisions model—placing dispositive weight on the absence of an express adherence pact, which Interstate Circuit never demanded even of human conspiracies, and requiring plaintiffs to come close to negating independence at the threshold—erroneously imports Matsushita’s evidentiary demand into Rule 12.22 The divergent outcomes that followed reflect two things at once: complaints that genuinely differed in what they alleged about data flows and participant knowledge, and courts that assigned sharply different inferential weight to common use of an intermediary. The first difference is legitimate; the second is the problem.
What the sustaining courts saw deserves a name. Consider what the use of a data-pooling algorithm entails. A firm that feeds its own nonpublic, transaction-level price and supply data into a shared pricing system has placed its most sensitive competitive intelligence into a tool that uses it to generate recommendations for known competitors. The firm receives something in return—forecasting and optimization services—but comparable analytics can be purchased without arming competitors. What ordinary commercial logic strains to explain is the further choice to let one’s own confidential data inform competitors’ prices. Standing alone, this is a dangerous act: a rival that knows a firm’s prices, discounts, and suppliers is well positioned to undercut it or otherwise seize advantage at its expense. The choice makes sense on one assumption: that one’s rivals are contributing too, in which case every participant is at once hostage and beneficiary of the arrangement, and none need fear being undercut. Twombly does not require that such conduct be inexplicable except as conspiracy.23 It requires that conspiracy be a plausible reading, and conduct this difficult to explain as unilateral—and thus suggestive of mutual assurance—supplies the requisite plausibility.24 That is the against-self-interest inference, a factor courts often describe as especially probative, specified for the algorithmic setting.25
The cases sustaining these claims already reason in these terms. In RealPage, the Middle District of Tennessee identified as the “most persuasive evidence of horizontal agreement” the “simple undisputed fact” that each landlord provided RealPage its proprietary commercial data “knowing that RealPage would require the same from its horizontal competitors and use all of that data to recommend rental prices to its competitors.”26 The complaint’s mechanics made the reciprocity concrete: RealPage maintained “peer lists” telling each client exactly whose nonpublic data would drive its pricing, and clients could review those lists, comment on them, and request that particular competitors be added.27 The Western District of Washington in Duffy sustained a claim on similar allegations, reasoning that entrusting Yardi with competitively sensitive data was not in any lessor’s independent self-interest absent an agreement,28 and finding “both invitation and acceptance” in the landlords’ understanding that general adherence to the recommendations was critical to the system’s success.29 The Third Circuit has now adopted the same reasoning precedentially, sustaining the claim where the software received “from each client non-public commercial information, and in return, giv[es] each client the benefit of their competitor’s non-public data.”30
What those courts reasoned case by case, this Contribution states as a rule: where a complaint plausibly alleges reciprocal contribution of current, nonpublic, competitively sensitive data to a common pricing system that uses those data in producing recommendations for known competitors, with each contributor aware of the reciprocal arrangement, it has alleged a plus factor ordinarily sufficient—joined with parallel conduct and the remaining elements of the claim—to support a plausible inference of agreement. The qualification about recommendations is critical; merely licensing a tool that also serves rivals is not enough. The operative allegation is that one participant’s information helps generate the recommendations for competing participants.
Reciprocal pooling is sufficient for the inference—but not necessary to it. The distinction matters more than it may appear. In MultiPlan, the Northern District of Illinois sustained a hub-and-spoke claim on a different footing: not data-pooling as such, but allegations that the intermediary communicated competitively sensitive positioning among payors—marketing its ability to bring one insurer “back into alignment with its primary competitor group”—and that payors agreed to abide by its rate calculations.31 The United States government has likewise pressed theories of concerted action that do not depend on data-pooling at all.32 If pooling were instead treated as a required element, complaints alleging concerted action on some other grounds—as in MultiPlan—would fail for want of pooling allegations, and the plus factor proposed here would instead become a barrier to recovery. This Contribution’s proposal is therefore deliberately narrow: pleading reciprocal data-pooling is one way to render an alleged algorithmic collusion agreement plausible, not a requirement that every complaint must satisfy.
Nor is the line an academic construction; it has been administered, and it cuts both ways. In Mach v. Yardi, a California court granted summary judgment for algorithm defendants in substantial part because discovery showed that the software did not use one customer’s nonpublic data to generate pricing recommendations for competitors.33 And when the United States moved to resolve its own suit against RealPage, the proposed consent judgment filed in November 2025—which RealPage agreed to follow pending the court’s Tunney Act review—drew the remedial line around substantially the same conduct: RealPage must stop using nonpublic competitor data in the runtime operation of its pricing products, but it remains free to train its models on data at least twelve months old.34 The consent decree’s conditions mark the boundary with the older information-exchange cases, which generally treat aggregation, anonymization, age, and independent third-party management as circumstances reducing, though not eliminating, competitive concern.35 Strip any condition—make the data public, or stale, or one-way, or remove the participants’ knowledge of one another—and the conduct recedes toward ordinary benchmarking, which the Sherman Act does not condemn.36
State legislatures have taken notice. California’s 2026 amendments to the Cartwright Act, its state antitrust statute, provide that a conspiracy complaint need not allege facts tending to exclude the possibility of independent action—Matsushita’s evidentiary demand—while separately prohibiting “common pricing algorithms” in terms that reach even public data.37 But the amendment is state antitrust law, and its application in federal court, where these class actions live, may generate an Erie question: federal pleading procedure ordinarily governs even as state law defines the substantive claim.38 New York legislatively deems specified participation in a coordinating rental-pricing system to constitute or facilitate an unlawful agreement, dispensing with ordinary plus factor proof, and cities including San Francisco have enacted rent-algorithm ordinances of their own.39 Statutes aimed at personalized “surveillance” pricing of individual consumers, such as Maryland’s 2026 act, address a different problem—discrimination across customers, not coordination among competitors—and the two should not be conflated.40 The result is a patchwork whose most calibrated instrument may not govern in federal court, and whose clearly applicable instruments are flat bans—assembled under a federal administration committed to challenging state AI regulation.41 State legislative action does not mean that courts should stand down; rather, it is evidence of what happens when courts misread the tools they already hold. The Sherman Act does not need amending to reach price fixing facilitated by an intermediary consultant, and it does not need amending to reach software that serves the same function.
The most serious objection to this Contribution’s proposed plus factor sounds in economics, and its core is eminently understandable. Contributing data to a shared algorithm, the argument runs, can be unilaterally rational: in return for its data, the contributing firm receives forecasts and analytics superior to anything it could build alone, so the disclosure of data reflects a purchase of forecasting power rather than an exchange of assurances. The Gibson district court credited a version of this reasoning, observing that a tool that improves over time “merely suggests [it] might be compelling . . . because it offers better pricing recommendations than it used to.”42 Most firms adopt these tools for unremarkable reasons, and nothing in the proposed factor assumes otherwise. But this objection confuses two different stages of a lawsuit. To defeat the factor at the pleading stage, it is not enough that an innocent explanation for data-sharing exists. Twombly asks only whether conspiracy is a plausible reading of the facts alleged, not the only reading or even the most likely one.43 The more difficult standard this objection contemplates—requiring evidence that “tends to exclude” innocent explanations—comes from Matsushita, and it applies at summary judgment, where discovery is completed and there is actual evidence to weigh.44 At the pleading stage, plausibility suffices,45 and a competing explanation is a matter for discovery and, where the rule of reason applies, merits analysis.
Economically, the objection proves less than it appears to. The unilateral-benefit account explains why a firm would subscribe to an aggregated, anonymized, or historical data service; that is the benchmarking the information-exchange cases already accommodate. It strains to find a pro-competitive explanation—or any plausible characterization other than tacit agreement—why a firm would hand granular, current, transaction-level data to a system that feeds it into recommendations for named rivals, under peer-list arrangements that make the reciprocity explicit. Better forecasts can be had without arming one’s competitors, and sellers of market analytics offer exactly that. What the pooled arrangement uniquely supplies is the assurance that rivals are pricing from the same data, an assurance most valuable to firms that intend to rely on it. The efficiency critique does not refute the proposed plus factor; it defines the factor’s boundary, and the four conditions—that the data is nonpublic, current, reciprocally used in rival recommendations, and knowingly pooled—are drawn where the critique runs out.
A second objection invokes Twombly’s own rationale: antitrust discovery is ruinously expensive, and a permissive gate invites suits that settle on nuisance value rather than merit.46 The concern is real, but it indicts the wrong rule. The factor proposed here is narrow and concrete—four conjunctive conditions, each pleadable only with specific facts about the system’s inputs and mechanics—not the mere parallelism Twombly feared. Moreover, Mach demonstrates from the defense side that where the facts do not bear the conditions out, defendants prevail.47 If anything, a defined plus factor is the more disciplined regime. Closely related pricing architectures have produced opposite outcomes depending on what complaints alleged about data sharing, knowledge, and adherence; the four conditions explain why those distinctions matter, and they provide materially clearer notice about which product features are most likely to support an inference of concerted action—the same features the proposed RealPage decree would prohibit.
The final objection takes its cue from Gibson: a shared adviser is not a conspiracy. The district court there embraced defense counsel’s analogy of an attorney who improves with each engagement, growing smarter from every client’s confidences—and who, as the court was careful to note, “does not share one client’s confidential information with another.”48 And antitrust law does not oblige a business to turn a blind eye to information its competitors also possess.49 Both propositions are correct, but neither is implicated. The court’s own caveat provides the analogy’s logical limit: the well-informed attorney’s clients receive her accumulated experience, not one another’s confidences. The analogy holds exactly where the factor’s conditions fail—where information flows one way, or is public, stale or anonymized—and under such circumstances, this Contribution’s approach does not meet it. An intermediary that pools rivals’ confidences and hands them back as prices is not the attorney contemplated by the Gibson court. That is not consultation; it is coordination.
Return to our consultant, who quietly collected rivals’ pricing data and told each firm what to charge, each firm knowing the others were in. From Interstate Circuit forward, courts have inferred agreement from what the participants did and knew—the structure of the arrangement itself—even where those participants never directly communicated with one another.50 Under the approach proposed by this Contribution, the algorithmic version of that consultant fares no differently: a complaint alleging that competitors knowingly pooled current, nonpublic data into a system that priced them all pleads a plausible agreement and proceeds to discovery. Under the dismissing courts’ approach, that same complaint fails unless the participants also, in effect, agreed to obey the consultant’s advice. Recognizing reciprocal data-pooling as a plus factor asks courts only to treat a concrete, recurring structure as the strong inference Twombly already allows; to refuse to let the substitution of code for conversation launder coordination the Sherman Act has reached since Interstate Circuit; and to stop demanding at the courthouse door the proof that discovery is designed to produce. The plus factor rests on one of the most probative inferences in the plus factor canon, it draws a line that recent practice proves is administrable from both sides, and it accomplishes judicially and uniformly what state legislatures are attempting piecemeal by statute. The appellate courts have begun to choose, and they have split: the Ninth Circuit, in Gibson, entrenched the error, and the Third Circuit, in Cornish-Adebiyi, supplied the correction. What remains is for other courts to follow—to name the inference, apply Twombly as written, and let the line between competition and collusion survive its translation into algorithms.
* Malcolm Girand is a J.D. Candidate (2027) at New York University School of Law. This Contribution is a commentary on the problem at the 2026 Antitrust Moot Court Competition hosted by the Antonin Scalia School of Law, George Mason University. One of the questions presented was whether common use of an algorithmic discounting tool constituted anticompetitive collusion. This Contribution distills one side of the argument, and the views expressed herein do not necessarily represent the author’s views.
1. Interstate Circuit, Inc. v. United States, 306 U.S. 208, 221, 226–27 (1939) (inferring horizontal agreement from knowing participation in a plan whose success depended on concerted action).
2. See, e.g., United States v. Socony-Vacuum Oil Co., 310 U.S. 150, 223 (“the machinery employed by a combination for price-fixing is immaterial”); Am. Tobacco Co. v. United States, 328 U.S. 781, 809–810 (1946) (antitrust conspiracy “may be found in a course of dealing or other circumstances as well as in an exchange of words”).
3. 15 U.S.C. § 1; Standard Oil Co. of New Jersey v. United States, 221 U.S. 1, 59–62 (1911) (construing statute to reach only undue or unreasonable restraints).
4. Bell Atl. Corp. v. Twombly, 550 U.S. 544, 553 (2007) (quoting Theatre Enters., Inc. v. Paramount Film Distrib. Corp., 346 U.S. 537, 540 (1954)).
5. See Theatre Enters., 346 U.S. at 540–41 (“[T]his Court has never held that proof of parallel business behavior . . . itself constitutes a Sherman Act offense.”); see also Twombly, 550 U.S. at 553–54.
6. Gibson v. Cendyn Grp., LLC, 148 F.4th 1069 (9th Cir. 2025), cert. denied, 224 L. Ed. 2d 502 (2026).
7. In re RealPage, Inc., Rental Software Antitrust Litig. (No. II), 709 F. Supp. 3d 478 (M.D. Tenn. 2023); Duffy v. Yardi Sys., Inc., 758 F. Supp. 3d 1283 (W.D. Wash. 2024); in re MultiPlan Health Ins. Provider Litig., 789 F. Supp. 3d 614 (N.D. Ill. 2025).
8. Cornish-Adebiyi v. Caesars Entm’t, Inc., No. 24-3006, 2026 WL 2182291, at *2, *10–11 (3d Cir. July 29, 2026), rev’g No. 1:23-cv-02536, 2024 WL 4356188 (D.N.J. Sept. 30, 2024). Both cases involved Cendyn’s Rainmaker platform. Compare id. at *2, with Gibson, 148 F.4th at 1079–80.
9. Twombly, 550 U.S. at 556–557. Under Federal Rule of Civil Procedure 12(b)(6), a defendant may move to dismiss a complaint that fails to state a claim. Twombly requires the complaint’s factual allegations to make the claim plausible, not merely conceivable, before discovery begins.
10. Twombly, 550 U.S. at 556–57; C-O-Two Fire Equip. Co. v. United States, 197 F.2d 489, 493 (9th Cir. 1952) (defining “plus factors”).
11. Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (plausibility “is not akin to a probability requirement”) (internal quotations omitted).
12. Monsanto Co. v. Spray-Rite Serv. Corp., 465 U.S. 752, 764 (1984); Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 588 (1986).
13. See, e.g., RealPage, 709 F. Supp. 3d at 502 (such conduct “will consistently tend to exclude the likelihood of independent conduct”) (quoting Re/Max Int’l, Inc. v. Realty One, Inc., 173 F.3d 995, 1009 (6th Cir. 1999)); in re Pool Prods. Distrib. Mkt. Antitrust Litig., 988 F. Supp. 2d 696, 711 (E.D. La. 2013) (presence of behavior otherwise inexplicable sans an anticompetitive agreement is “generally considered the most important plus factor”) (internal quotations omitted).
14. Compare RealPage, 709 F. Supp. 3d at 519–20 (declining per se treatment), with Duffy, 758 F. Supp. 3d at 1296 (disagreeing).
15. Compare United States v. Container Corp. of Am., 393 U.S. 333, 337 (1969) (condemning exchange of current, customer-specific prices), and United States v. U.S. Gypsum Co., 438 U.S. 422, 441 n.16 (1978) (“Exchanges of current price information, of course, have the greatest potential for generating anti-competitive effects . . . . ”), with Maple Flooring Mfrs. Ass’n v. United States, 268 U.S. 563, 586 (1925) (practice of aggregating historical trade-association data upheld). See also Todd v. Exxon Corp., 275 F.3d 191, 211–12 (2d Cir. 2001) (weighing data currency and specificity and finding the practice troubling).
16. Gibson, 148 F.4th at 1076.
17. Id. at 1084.
18. Id. at 1083 n.8 (“This analysis might change if Plaintiffs had alleged that Cendyn shared the confidential information of each competing hotel among the licensees.”).
19. See Cornish-Adebiyi, 2024 WL 4356188, at *5–7 & n.4 (dismissing where complaint did not plausibly allege pooling or exchange of confidential data).
20. Gibson v. MGM Resorts Int’l (Gibson I), No. 2:23-cv-00140, 2023 WL 7025996, at *5 (D. Nev. Oct. 24, 2023), quoted in RealPage, 709 F. Supp. 3d at 512.
21. Cornish-Adebiyi, 2026 WL 2182291, at *11.
22. See supra notes 1 and 12 and accompanying text.
23. See Twombly, 550 U.S. at 556–57.
24. Id. at 556–557.
25. See, e.g., Interstate Circuit, 306 U.S. at 222 (“Each was aware that all were in active competition and that without substantially unanimous action[,] . . . there was risk of a substantial loss.”); Toys “R” Us, Inc. v. FTC, 221 F.3d 928, 935–36 (7th Cir. 2000) (“[T]he only condition on which each toy manufacturer would agree . . . was if it could be sure its competitors were doing the same thing.”).
26. RealPage, 709 F. Supp. 3d at 510.
27. Id. at 512–13 (quoting the Multifamily Complaint ¶ 289).
28. Duffy, 758 F. Supp. 3d at 1293–94.
29. Id. at 1294.
30. Cornish-Adebiyi, 2026 WL 2182291, at *11.
31. MultiPlan, 789 F. Supp. 3d at 641–42 (payors agreed “to abide by MultiPlan’s rate calculations,” and MultiPlan promoted “alignment” with competitors).
32. See, e.g., Statement of Interest of the United States at 2–3, Duffy, No. 2:23-cv-01391 (W.D. Wash. 2024), ECF No. 149 (arguing that competitors’ “jointly delegating key aspects of their decisionmaking to a common algorithm” is per se unlawful even where participants “retain some authority to deviate”); Statement of Interest of the United States at 4–6, MultiPlan, No. 1:24-cv-06795 (N.D. Ill. 2025), ECF No. 382 (arguing that joint use of a common pricing algorithm may constitute concerted action “even if the competitors do not always use the algorithm the same way”).
33. Mach v. Yardi Sys., Inc., No. 24-CV-063117 (Cal. Super. Ct., Alameda Cnty. Oct. 2025).
34. Proposed Final Judgment, United States v. RealPage, Inc., No. 1:24-cv-00710 (M.D.N.C. Nov. 24, 2025) (stipulation requires interim compliance; entry would have no prima facie effect in private litigation).
35. See supra note 15 and accompanying text.
36. Maple Flooring, 268 U.S. at 585–586.
37. Cal. Bus. & Prof. Code § 16756.1 (effective Jan. 1, 2026).
38. See Hanna v. Plumer, 380 U.S. 460, 465, 471 (1965); Shady Grove Orthopedic Assocs., P.A. v. Allstate Ins. Co., 559 U.S. 393, 398–99 (2010) (valid federal rule governs in federal court notwithstanding contrary state law).
39. N.Y. Gen. Bus. Law § 340-b (effective Dec. 15, 2025); see also S.F., Cal., Rent Ordinance Sec. 37.10C (2024).
40. Md. H.B. 895, 2026 Leg., Reg. Sess. (signed Apr. 28, 2026; effective Oct. 1, 2026); N.Y. Gen. Bus. Law § 349-a.
41. See Exec. Order 14365, Ensuring A National Policy Framework for Artificial Intelligence, (Dec. 2025) (“United States AI companies must be free to innovate without cumbersome regulation. But excessive State regulation thwarts this imperative.”); White House, National Policy Framework for Artificial Intelligence (Mar. 2026).
42. See Gibson v. Cendyn Grp., LLC (Gibson II), No. 2:23-cv-00140, 2024 WL 2060260, at *6 (D. Nev. May 8, 2024); see also Brief of Int’l Ctr. for Law & Econ. as Amicus Curiae Supporting Defendants-Appellees, Gibson, No. 24-3576, at *15 (9th Cir. 2024) (warning that condemning shared pricing tools would deter innovation).
43. Twombly, 550 U.S. at 556.
44. Matsushita, 475 U.S. at 588.
45. Iqbal, 556 U.S. at 678.
46. See Twombly, 550 U.S. at 558–59 (“The threat of discovery expense will push cost-conscious defendants to settle even anemic cases.”).
47. See Mach, No. 24-CV-063117, slip op. at 8–9 (granting summary judgment because “the record lacks any evidence of the alleged unlawful ‘give-to-get’ scheme” and “there is no rim”).
48. Gibson II, 2024 WL 2060260, at *6.
49. Gibson, 148 F.4th at 1084.
50. Interstate Circuit, 306 U.S. at 221, 226–27.