Farella Quarterly Ninth Circuit Antitrust Review - Q3 2026

In a significant decision for refusal-to-deal claims under Section 2 of the Sherman Act, the Ninth Circuit reversed summary judgment in Simon and Simon, PC v. Align Tech., Inc., 2026 WL 2980631 (9th Cir. Oct. 5, 2026), and held that a defendant’s legitimate business justification does not end the refusal-to-deal inquiry. Under the clarified framework, a plaintiff may still prevail on a refusal-to-deal claim by showing that the defendant’s justification is pretextual or not legitimately procompetitive, or that the conduct’s anticompetitive harm outweighs its benefits. The decision brings rule-of-reason balancing into the Ninth Circuit’s refusal-to-deal analysis and aligns this circuit with the Seventh Circuit.

In Align, two groups of plaintiffs sued Align Technology, Inc., the maker of the Invisalign clear aligner and the iTero intraoral scanner, claiming that Align violated Section 2 of the Sherman Act by scheming to monopolize the aligner and scanner markets. During the relevant period, Align held about 90% of the U.S. aligner market and had enabled digital interoperability between Invisalign and a few competing scanners. In December 2015, Align agreed to make 3Shape’s TRIOS scanner one of the scanners that is compatible with Invisalign. Align repeatedly asked 3Shape to make TRIOS work only with Invisalign, and 3Shape refused. In December 2017, with a year left on the agreement, Align announced that it would “no longer be able to accept digital scans for new Invisalign treatment and/or retention cases from TRIOS scanners in the United States.”  Align, 2026 WL 2980631, at *3. This came at a time when key Invisalign patents were about to expire and TRIOS had become a growing channel for orders. About a month earlier, Align had filed six patent actions against 3Shape. Meanwhile, Align kept its interoperability deals with 3M and Dentsply Sirona, and kept accepting TRIOS scans outside the United States. The plaintiffs claimed this refusal to deal was part of a broader scheme that also bundled iTero and Invisalign sales.

Judge Chhabria of the Northern District of California granted summary judgment to Align. Relying on the ABA model jury instructions, the court held that “[a] refusal to deal that is based in part on legitimate business reasons does not violate the antitrust laws, even if it is also motivated by the desire to harm competitors.”  Align, 2026 WL 2980631, at *4. Align argued that continuing interoperability would have weakened its position in the patent litigation by allowing 3Shape to raise various equitable defenses. The court found that Align’s patent enforcement was a “presumptively valid business justification,” and concluded that no reasonable jury could reject the idea that the termination was based “in part” on legitimate reasons. Id. at *10. The court did not weigh the conduct’s anticompetitive harm against its benefits. 

The Ninth Circuit reversed and remanded in a per curiam opinion, calling the district court’s analysis “an incomplete application” (Align, 2026 WL 2980631, at *2) of the burden-shifting framework, and clarified that framework in three steps: 

1. Prima facie case. The plaintiff must show anticompetitive conduct, and the factors from Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985) are applied flexibly. All three factors were met: Align ended a voluntary and profitable course of dealing, abandoned retail-price sales, and singled out 3Shape while selling through other scanners. 
2. Justification. The defendant must offer a procompetitive, nonpretextual justification. Align met this burden by pointing to its patent litigation. 
3. Rebuttal or balancing. The plaintiff may show the justification is not legitimately procompetitive or is pretextual. If it cannot, it may still win by showing the harm outweighs the benefit. 

On the merits, plaintiffs’ patent expert testified that ending interoperability actually weakened Align’s patent position. That created a triable issue on whether the justification was legitimately procompetitive, while internal documents and the timing of the litigation also supported a finding of pretext. The most significant outcome of this holding is that the rule of reason, including balancing harms against benefits, is now part of the Ninth Circuit’s refusal-to-deal framework. The panel described Fed. Trade Comm’n v. Qualcomm Inc., 969 F.3d 974, 991 (9th Cir. 2020), as having implied this rule and said it was joining the Seventh Circuit’s approach in Viamedia, Inc. v. Comcast Corp., 951 F.3d 429, 462 (7th Cir. 2020). Several other points are significant for future refusal to deal cases:

• Higher bar for justifications. A justification must be “a real one”—not just “conceivable”—which ends the broad reading of Aerotec Int’l, Inc. v. Honeywell Int’l, Inc., 836 F.3d 1171, 1184 (9th Cir. 2016).
• Legitimacy is a separate question from pretext. Even a sincere justification can be challenged on whether it actually promotes competition. 
• The deal is not the product. Interoperability is a way to deliver the product, not the product itself, so a defendant cannot avoid Aspen Skiing by saying it never sold “interoperability.”
• No minimum “course of dealing.”  A short-term, terminable-at-will agreement can still count. 

Judge Miller concurred only in the judgment. He agreed there was a triable question regarding pretext, but he called the balancing holding “a significant and unwarranted expansion of refusal-to-deal liability.”  Align, 2026 WL 2980631, at *14. In his view, the Supreme Court has never required balancing in refusal-to-deal cases. He described Qualcomm’s language as “unreasoned dicta” and warned that a “freewheeling balancing rule” asks courts to act as central planners and risks chilling procompetitive conduct. Id.

While it remains to be seen how Judge Miller’s concerns will play out in the lower courts, the Ninth Circuit’s decision will make refusal-to-deal claims harder to dispose of on summary judgment. Defendants can no longer rely on a plausible business reason alone; they should be prepared to show that the justification is real, that it actually promotes competition, and that its benefits outweigh any harm to competition. Plaintiffs, in turn, gain a clearer path to trial by challenging the legitimacy of a justification or proving net anticompetitive effects.

In Blythe v. Nat’l Collegiate Athletic Ass’n, 190 F.4th 857 (9th Cir. 2026), the Ninth Circuit joined the Third, Fourth, and Seventh Circuits in holding that the NCAA’s “Five-Year Rule” is a commercial restraint subject to Section 1 of the Sherman Act, while simultaneously confirming that college athletes must build a current, sport-specific evidentiary record about the relevant market and market-wide anticompetitive effects to win relief—something plaintiff failed to do.

NCAA Bylaw 12.6 limits athletes to four seasons of intercollegiate competition in a sport, which must be completed within five calendar years of first full-time enrollment at any “collegiate institution,” a term that includes two-year, non-NCAA, and foreign schools. Plaintiff Noah Blythe first enrolled and played baseball at the Division II Hawaii Pacific University in 2020, then played for three more seasons (2021–2023) at the University of Antelope Valley until it closed (affiliated with the National Association of Intercollegiate Athletics, rather than the NCAA) and returned to Hawaii Pacific for a fifth year in 2024. Hoping for extra innings, he then transferred to the University of Nevada, Reno, a Division I program, which sought a waiver of the Five-Year Rule so he could play the 2026 season. Both the conference and the NCAA denied the request. Blythe sued in the District of Nevada, alleging that the rule violates Section 1, and the district court (Traum, J.) granted a preliminary injunction after defining the relevant market as the “Division I collegiate baseball labor market.”  On the NCAA’s expedited appeal (the appellate version of instant replay), the Ninth Circuit vacated the injunction on April 8, 2026, while Blythe was playing under it, and issued this opinion (Sanchez, J., joined by Ezra, D.J.; Bennett, J., concurring) explaining its decision on September 9.

 The panel held, as a matter of first impression in the circuit, that the Five-Year Rule is a commercial restraint. Its opinion built on the “sea change” of Alston, 594 U.S. 69 (2021), in which the Supreme Court held that the NCAA’s educational compensation rules were not noncommercial amateurism rules and were subject to the Sherman Act. Relying on Alston’s refusal to grant the NCAA “judicially ordained immunity” for rules sitting “at the intersection of higher education, sports, and money” (Blythe, 190 F.4th at 862), the court reasoned that eligibility rules affecting commerce are not exempt from antitrust scrutiny. It expressly adopted the reasoning of Elad v. Nat’l Collegiate Athletic Ass’n, 160 F.4th 407, 415 (3d Cir. 2025); Fourqurean v. Nat'l Collegiate Athletic Ass’n, 143 F.4th 859, 868 (7th Cir. 2025); and Robinson v. Nat’l Collegiate Athletic Ass’n, 172 F.4th 271, 289 (4th Cir. 2026), each of which treated the rule as a restraint on athletes’ participation in a labor market. In a footnote, the court rejected the NCAA’s reliance on the suggestion in O’Bannon v. Nat’l Collegiate Athletic Ass’n, 802 F.3d 1049 (9th Cir. 2015) that rules “limiting the number of years that student-athletes may play” are noncommercial “true eligibility” rules, characterizing that passage as non-binding dicta. Blythe, 190 F.4th at 861 n.3.

The panel nonetheless held that Blythe was out at first base—that is, plaintiff had offered almost no evidence at step one of the three-part rule-of-reason framework (requiring proof that the challenged restraint has a substantial anticompetitive effect that harms consumers in the relevant market). One problem was market definition: the district court erred by borrowing Alston’s market, a market limited to basketball and football that did not address baseball. The reliance on Alston’s 2021 analysis was further misplaced because college sports have changed “dramatically” in the past five years with the advent of athlete compensation for performance and name, image, and likeness (NIL) deals. Echoing its sister circuits, the court stressed that reliance on “a previously accepted market, without inspection of current market realities, is antithetical to antitrust legal principles.”  Blythe, 190 F.4th at 865. Blythe “presented almost no concrete evidence to define the alleged collegiate baseball market, let alone its changes post-Alston.”  Id. His record (a generalized report about the collegiate NIL market, a statistic that 71% of 2025 MLB draft picks came from Division I programs, and his own declaration that he received no NIL offers at the NAIA and Division II levels), and the district court’s “cursory” analysis, said nothing about market participants, dynamics, or which employers, such as professional or other college leagues, athletes view as reasonable substitutes.

The court also rejected Blythe’s argument that the rule’s horizontal character relaxed the market definition requirement, reasoning that the Supreme Court has loosened market definition requirements only for agreements among competitors not to compete, and Alston demands careful market analysis even for admitted horizontal price fixing. Finally, even assuming a properly defined market, Blythe showed no substantial anticompetitive effects. He offered only his own exclusion, not evidence of any injury to competition (i.e., adverse effects on NCAA rivals). He presented no other evidence that the Five-Year Rule depressed compensation or caused other anticompetitive effects. Quoting Fourqurean, the court further observed that a restraint limiting labor supply would ordinarily increase, not decrease, worker compensation. Blythe, 190 F.4th at 864. Having found no likelihood of success, Blythe thus failed to make the required showing for a preliminary injunction. Judge Bennett joined the majority’s rule-of-reason analysis and agreed that the injunction should be vacated, but he would not have decided whether the Five-Year Rule is commercial. 

After Blythe, a four-circuit consensus now exists in two respects, with a decision from the Tenth Circuit pending. First, after Alston, NCAA eligibility rules are not categorically shielded from Section 1 liability, and the NCAA’s efforts to treat such rules as noncommercial or as product features, including through O’Bannon’s dicta, have lost traction. Second, despite this treatment, plaintiffs have gone 0-for-4: the Third, Fourth, Seventh, and now Ninth Circuits have each vacated preliminary injunctions against the Five-Year Rule or related rules for failure to prove a relevant market and market-wide effects. But the Ninth Circuit did not suggest that the Five-Year Rule will always withstand scrutiny. This case was not a close call: the court’s opinion calls out both the plaintiff and the court for superficial analysis unsupported by an evidentiary record, and the attempt to suggest that the market for college baseball is the same as football or basketball is obviously erroneous. A plaintiff with better developed economic evidence, including a sport-specific labor market reflecting post-Alston NIL and compensation realities, analysis of substitutes (Division II and III, NAIA, junior colleges, professional and foreign leagues), a theory of harm to competition rather than to the individual athlete, and an answer to the supply-restriction argument, could fare better. For the NCAA and member institutions, the decision supplies a playbook for opposing preliminary relief at step one, though the commerciality holding ensures that eligibility rules will face full rule-of-reason scrutiny on a developed record. Practitioners should also expect Alston’s statement that legal analysis may change with “market realities” to cut both ways as the economics of college sports continue to evolve. Alston, 594 U.S. at 102.

The Ninth Circuit’s decision in Surgical Instrument Serv. Co., Inc. v. Intuitive Surgical, Inc., 187 F.4th 1092 (9th Cir. 2026), clarified the requirements for a single-brand aftermarket in light of the competitive conditions in the foremarket.

Defendant Intuitive Surgical, Inc. (“Intuitive”) manufactures and sells the “da Vinci” surgical robot used to perform minimally invasive soft-tissue (“MIST”) surgery. Intuitive possesses more than 99% market share in the market for MIST surgical robots—the foremarket—and 100% share in the robot attachment instruments used in such surgeries—the “aftermarket.”  Plaintiff Surgical Instrument Service Company, Inc. (“SIS”) is a surgical instrument repair business that challenged Intuitive’s practice of limiting the number of pre-set uses of Intuitive’s “EndoWrist” surgical instrument. SIS asserted claims for tying, exclusive dealing, monopolization, and attempted monopolization in the single-brand aftermarket for repair and replacement of Intuitive’s MIST surgical robot instruments.

At trial, the district court instructed the jury that, to establish such a single-brand aftermarket, SIS must prove each of the four Kodak/Epic factors articulated in the Ninth Circuit’s decision in Epic Games, Inc. v. Apple, Inc., 67 F.4th 946 (9th Cir. 2023): (1) the challenged restrictions are generally unknown; (2) significant information costs prevent forecasting life-cycle pricing; (3) there are significant switching costs; and (4) and general market-definition principles regarding the cross-elasticity of demand. Acknowledging that it had presented no evidence at trial to establish these factors, SIS stipulated to judgment in Intuitive’s favor and appealed. 

The Ninth Circuit reversed in an opinion by Judge Koh. The Ninth Circuit held that “the Kodak/Epic factors apply only in cases where the plaintiff cannot show that the defendant possesses market power in the competitive foremarket” (Surgical Instrument,187 F.4th at 1108)—the special case of a “Kodak-style antitrust claim.”  Id. at 1103. The purpose of the factors is to determine “whether the competition and information available in the foremarket is sufficient to discipline anticompetitive conduct in the aftermarket.”  Id. at 1109. If it is not, then the defendant may possess market power in the aftermarket even where it lacks market power in the foremarket—as was the case in Eastman Kodak Co. v. Image Tech. Servs. Inc., 504 U.S. 451 (1992). But where there is no meaningful competition in the foremarket, it is not necessary to determine whether, for example, the challenged aftermarket restrictions are generally known; such knowledge would not “impact consumers’ purchasing decisions in the foremarket, thereby constraining the defendant’s ability to raise prices in the aftermarket.”  Surgical Instrument,187 F.4th at 1110. 

The rules for pleading and proving a single-brand aftermarket in the Ninth Circuit are now clear. “[A] plaintiff seeking to pursue an antitrust claim based on the defendant’s possession of market power in a single-brand aftermarket has two options: (1) show that the defendant has market power in the foremarket, or (2) if the plaintiff cannot show that the defendant possesses market power in the foremarket, satisfy the Kodak/Epic factors.”  Id. at 1110.

The Surgical Instrument decision is notable in three respects. First, clarification regarding single-brand aftermarkets is important given the increasing prevalence of the “closed ecosystem” model. Indeed, these markets recently have been litigated in cases involving both technology platforms (e.g., Epic) and the “right to repair” (e.g., John Deere, Tesla). Second, the Ninth Circuit was not persuaded by Defendant’s assertion that single-brand aftermarkets are “disfavored” because they “risk making every company a monopolist in its own brand.”  Surgical Instrument, 187 F.4th at 1109 n.6. The court noted that this concern was not present where the defendant already possessed power in the foremarket and used that power to require purchases in the aftermarket. Third, the decision cements Judge Koh’s status as a leading antitrust jurist on the Ninth Circuit. It also exemplifies the lack of excessive formalism in her antitrust jurisprudence and her willingness to focus on the underlying competitive concerns—rather than any doctrinal structure.

The Northern District of California’s decision in Bryant v. Microsoft Corp., --- F. Supp. 3d ----, 2026 WL 2908104 (N.D. Cal. Sept. 28, 2026), demonstrates the difficulties in pleading a sufficient federal antitrust injury in a downstream market and also highlights the broader scope of California’s Cartwright Act.

Plaintiffs are a proposed class of ChatGPT subscribers who allege that an exclusivity agreement between Defendant Microsoft Corporation (“Microsoft”) and OpenAI injured end users of U.S. Consumer Generative AI (“CGAI”) applications. Beginning in 2019, Microsoft invested $1 billion in OpenAI in exchange for OpenAI agreeing to exclusively use Microsoft’s Azure cloud platform for processing capacity, also known as “compute.”  OpenAI needs compute for its ChatGPT models to train, run, and generate answers to users’ queries. Plaintiffs allege that the constraint of an exclusivity requirement raised OpenAI’s costs and limited OpenAI’s capacity, resulting in higher prices and lower quality for consumers in the U.S. CGAI market (the relevant market proposed by plaintiffs)—until OpenAI reached a compute agreement with Google and slashed prices by 80%. Plaintiffs assert claims under Section 1 of the Sherman Act, California’s Cartwright Act, and California’s Unfair Competition Law (“UCL”).

Microsoft moved to dismiss plaintiffs’ Sherman Act claim for failure to plead antitrust injury, arguing that plaintiffs’ injury occurred in the downstream CGAI market rather than the upstream compute market where the exclusivity agreement restraint operated. Addressing only whether plaintiffs had satisfied the requirements of antitrust injury, Judge Pitts dismissed the Sherman Act claim with leave to amend.

The court applied the Ninth Circuit’s four-element antitrust injury test and determined that plaintiffs had not satisfied the fourth element—injury of the type the antitrust laws were intended to prevent. Noting the Ninth Circuit’s general “corollary[ ] that the injured party [must] be a participant in the same market as the alleged malefactors,” the court reasoned that the restraint imposed by OpenAI’s exclusivity agreement with Microsoft did not operate in the CGAI market, and instead operated in the upstream compute market, which “[p]laintiffs do not participate in . . . at all.”  Bryant, 2026 WL 2908104, at *5-6.

The court suggested that in such circumstances—when a court’s evaluation of the fourth antitrust injury element reveals that the consequences of an antitrust violation were felt outside the immediate market in which the violation took place—courts in the Ninth Circuit may look to the factors considered by the U.S. Supreme Court in Associated Gen. Contractors of Cal., Inc. v. Cal. State Council of Carpenters, 459 U.S. 519 (1983): whether the plaintiff participated in the purportedly restrained market; the indirectness of the injury to the plaintiff; the speculative nature of the plaintiff’s damages; and the risk of duplicative or difficult-to-apportion damages.

The court determined that the Associated General Contractors factors weighed against plaintiffs satisfying the fourth antitrust injury element and being able to pursue a claim under Section 1 of the Sherman Act. Skipping the redundant first factor (market participation), which the court had already considered and turned to Associated General Contractors as a result, the court noted that plaintiffs had admitted the injury to ChatGPT subscribers was indirect, whereas OpenAI had suffered direct harm. As a result, “determining the extent to which the agreement—rather than, for example, OpenAI’s own market dominance—resulted in plaintiffs paying a higher price is somewhat speculative and may require apportioning damages as between plaintiffs and OpenAI itself.”  Bryant, 2026 WL 2908104, at *6.

The court then returned to the fundamental purpose of its inquiry: whether the injury alleged was of the type the antitrust laws were intended to prevent. The court concluded that the “core problem with plaintiffs’ theory of antitrust injury is that they allege harm to a competitor in the market for CGAI, rather than harm to the market itself.”  Id. According to the court, the constraint on OpenAI in the upstream compute market did not materially change the structure of the CGAI market, such as by changing the number of competitors or by changing how competitors could compete against one another. 

Microsoft also moved to dismiss plaintiffs’ California claims on the grounds that, as relevant here, (1) plaintiffs did not reside in California; (2) the Cartwright Act mirrors federal antitrust law, under which plaintiffs’ claims had failed; and (3) with respect to the UCL claim, plaintiffs had not shown that Microsoft caused them economic injury and plaintiffs had failed to allege that Microsoft’s conduct was either unlawful or unfair.

Plaintiffs’ state-law claims survived. First, the court concluded that even if plaintiffs resided outside California, they had pleaded enough conduct in California to overcome the UCL’s presumption against extraterritoriality. OpenAI is a San Francisco-based company, and plaintiffs had alleged that Microsoft’s AI division worked directly with OpenAI’s California-based engineers. Second, although the Cartwright Act requires antitrust injury, the court noted that California has expanded the scope of injured persons who may pursue antitrust claims to those who dealt directly “or indirectly” with the violator. Third, the court concluded that Microsoft’s arguments did not warrant dismissal of the UCL claim. Plaintiffs alleged that Microsoft’s conduct led to their economic injuries (albeit indirectly), and a challenge to a UCL unfairness claim calls for a balancing test that would be inappropriate at the pleadings stage. 

The Bryant decision stands out for three reasons. First, Bryant leaves open the possibility that a downstream injury could satisfy the Ninth Circuit’s antitrust injury test if the facts presented sufficiently contrasted with those in Associated General Contractors. Second, Bryant highlights the Cartwright Act as a meaningful alternative for indirect purchasers whose claims are too indirect for the Sherman Act. Third, Bryant suggests that the Northern District of California, home to the global capital of AI, may become a key venue for AI-related antitrust actions.