U.S. District Court for the District of Columbia | Nos. 25-cv-00543 and 25-cv-03192 | September 30, 2026

J.R. Howell  |  The American Counsel 

On September 30, 2026, Judge Amit P. Mehta dismissed the amended complaints in Chegg, Inc. v. Google LLC and Penske Media Corporation v. Google LLC. The publishers alleged that Google used its dominance in search to obtain their content without payment and then used that content to answer questions that otherwise would bring readers to their websites. The decision matters because it identifies several barriers between dependence on a dominant platform and a private antitrust remedy. Each barrier concerns a different legal question, including whether the parties agreed to exchange anything, whether the proposed markets make economic sense, and whether these plaintiffs can enforce competition in general search.[1]

The cases expose a problem that deserves attention even where the complaints fall short. A company may depend on a platform to reach its customers while supplying the material that helps the platform compete against it. If courts divide that relationship into separate markets, the company can experience the alleged exercise of monopoly power directly yet struggle to qualify as an antitrust plaintiff in the market where that power resides. A serious response requires a precise account of the competitive harm. It also requires courts to examine how the power operates across those market boundaries.

The reciprocal-dealing claims required an agreement the publishers had not pleaded

Chegg sells educational services. Penske and its affiliated publishers produce a range of media content, including publications such as Rolling Stone, Variety, and The Hollywood Reporter. As the opinion recounts their allegations, they depend on visitors who arrive through search to generate subscriptions, advertising revenue, and affiliate commissions. They described an exchange in which publishers provide material for Google’s search index and Google returns referral traffic.[2]

According to the complaints, that relationship changed as Google began displaying more of the answer itself. Featured snippets reproduced information on the search results page. AI Overviews could generate longer answers from publishers’ content. A reader who found an adequate answer on Google would have less reason to visit the source. The publishers further alleged that blocking the relevant uses of their content would require sacrificing their visibility in search. At this stage, these descriptions were allegations the court accepted for purposes of evaluating the pleadings.[3]

This account identifies the practical source of the dispute. While a publisher may technically withhold material from a crawler, the economic consequences of disappearing from search can make that option costly. The alleged power lies in controlling access to readers while also deciding how the work used to reach them will be repurposed. However, describing that dependence does not establish each element of the antitrust theories the publishers selected.

Their reciprocal-dealing claims encountered that difficulty first. They alleged that Google conditioned the provision of search traffic on receiving content for republishing, training AI models, and generating grounded answers. The court held that they had failed to plead an agreement. Neither an expectation of traffic nor a longstanding pattern of allowing crawling established mutual assent to a transaction. The complaints identified no negotiated terms or communications from which the required arrangement could plausibly be inferred.[4]

The court’s reasoning under Section 2 requires particular care. Section 2 reaches unilateral monopolizing conduct. Mehta did not hold that every Section 2 claim requires an agreement. He held that the specific practice these plaintiffs alleged, reciprocal dealing, presupposes a dealing relationship that they had not plausibly described. Their choice of theory therefore carried a requirement that simply invoking unilateral monopoly power could not avoid.[5]

Google’s strongest answer follows from that distinction. An expectation that search will continue sending visitors cannot by itself oblige a search engine to preserve a particular design or revenue stream. If indexing every publicly accessible website created a reciprocal contract, the court reasoned, Google would have agreements with billions of publishers. A persuasive challenge must identify the conduct that excludes competition with greater precision than an asserted entitlement to the earlier arrangement.

A useful answer can still raise a separate question about control

Penske also alleged that Google unlawfully tied AI Overviews to general search. The court rejected that claim because Penske had not plausibly alleged separate consumer demand for the two products. The fact that traditional search sends users to other websites while an AI summary may keep them on Google did not establish that they were distinct products for tying purposes.[6]

Indeed, the court found a tension within Penske’s own account. Its allegation that users accepted the AI answer rather than following links suggested a common demand for information. The complaint did not supply facts showing that users would choose Google for search while choosing a different provider to supply the accompanying AI summary. Mehta also considered the allegation that competing search engines offered comparable summaries.

There is force to this reasoning. Product improvement can reduce demand for another company’s offering without unlawfully restraining competition. Courts should not infer an illegal tie whenever a platform adds a feature that makes another business less useful. The separate-products inquiry calls for evidence about demand.

However, satisfaction with an answer and control over its supply remain analytically different questions. Evidence that users consume a bundled feature does not necessarily establish what they would choose if competing features were meaningfully available. That is my criticism of relying too heavily on observed use within the arrangement being challenged. It is also a research question that the complaints needed to answer with facts about available choices, user preferences, and competitive alternatives. The opinion found those facts missing. It did not establish that integration can never conceal exclusion.

Antitrust standing asks who may enforce the competition claim

The monopoly-maintenance claims failed on antitrust standing. This is distinct from Article III standing, which concerns injury in fact, traceability, and redressability. As the D.C. Circuit explained in Johnson v. Commission on Presidential Debates, antitrust standing asks whether a plaintiff is a proper party to pursue the particular private antitrust action. The September 30 opinion’s standing analysis concerns that claim-specific requirement. It does not hold that the publishers’ alleged financial losses fail Article III’s injury requirement.[7]

The distinction reflects the rule in Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc.. There, the Supreme Court rejected a damages theory based on profits the plaintiffs would have earned if competing bowling centers had closed. Keeping those rivals in business could injure the plaintiffs financially while preserving competition. Antitrust injury must therefore connect the claimed loss to the feature of the defendant’s conduct that makes it anticompetitive.[8]

The publishers attempted that connection in two ways. First, they described themselves as suppliers of search-index data and purchasers of referral traffic. Mehta concluded that their lost subscriptions and uncompensated content concerned publishing and content markets rather than general search. Their purchaser characterization also encountered the earlier failure to allege an agreement to purchase traffic.[9]

Second, they argued that their injuries were integral to Google’s exclusion of competing AI services. In their account, Google obtained content without payment while rivals had to license it. The publishers’ lost compensation therefore supplied the mechanism of an advantage in search. This argument had to establish more than the existence of different input costs.

The missing link was between free content and exclusion

The publishers relied on Blue Shield of Virginia v. McCready, in which a health-plan subscriber could pursue an antitrust action after being denied reimbursement for treatment by a psychologist. The alleged scheme used reimbursement rules to favor psychiatrists. The subscriber’s financial injury was part of the means by which the defendants allegedly sought to restrict competition. Her ability to sue did not depend on being a competing provider.[10]

Mehta found that the publishers had not pleaded a comparable connection. Google had acquired substantial search dominance before the recent introduction of AI Overviews. The complaints did not explain how obtaining content without payment distinguished Google from the search engines that historically crawled websites on the same basis. As to maintaining that dominance through newer AI products, the court wanted facts showing how the alleged cost advantage impaired competition. The complaints did not allege that all or most relevant rivals paid for training content or that those costs were prohibitive and produced anticompetitive effects.[11]

That is a consequential demand for precision. A publisher may be able to show what its content is worth through its licensing agreements. Those agreements alone do not establish the effect of Google’s access on entry or competition in general search. The missing questions concern which material rivals need, whether they can obtain substitutes, how access affects their costs, and whether the resulting disadvantage protects Google from competitive pressure.

At the same time, the inquiry should remain sensitive to the role of a supplier’s injury. A platform can potentially impose an uncompensated input requirement through control of distribution. If that requirement is part of a demonstrated mechanism for excluding rivals, the supplier’s loss may have a closer relationship to the restraint than a simple division between search and publishing suggests. That is an argument for examining the mechanism under the governing doctrine. It is not a finding that these plaintiffs supplied the missing allegations.

Private enforcement makes this distinction especially important. Businesses that bear an alleged restraint may possess knowledge and incentives that differ from those of a public agency. I favor preserving a meaningful route for them to challenge unlawful conduct. A doctrine that regards them as outsiders must give careful attention to whether their injury is an instrument of the restraint, even where their primary business occupies an adjacent market.

Market definition must describe substitution rather than the internet

The publishers also tried to challenge Google’s expansion into their own markets through monopoly-leveraging and attempted-monopolization claims. Here the opinion identified a different defect. Penske’s proposed online-publishing market covered essentially all text published online. It did not explain why a legal brief, a work of fiction, an educational answer, and a news article were substitutes. Their common digital format could not perform the work of a market definition.[12]

Chegg offered a narrower educational-publishing market. Nevertheless, the court found its boundaries too uncertain. Characteristics such as curation and verification also describe other nonfiction publishing. Meanwhile, Chegg’s allegations about how Google generated AI answers did not establish that Google participated in a market defined by those characteristics. The court therefore found no plausible basis for the asserted dangerous probability that Google would monopolize that market.[13]

This part of the decision identifies a discipline that platform cases need. To translate a company’s broad influence over information into an antitrust claim, the pleading must explain which offerings constrain one another and why. It must also connect the challenged conduct to power within that market.

Yet access to distribution deserves a place in that analysis. The court questioned how Google could restrict entry into a market where anyone could publish text online. In my view, the ability to upload material is only the beginning of the competitive inquiry. If commercially meaningful access to readers depends on a dominant intermediary, formal ease of publication may coexist with substantial obstacles to reaching an audience. Establishing that proposition would require evidence about substitutes for search referrals and the cost of building viable distribution. It cannot cure an overbroad market simply by assertion.

An enforcement response must reach the source of dependence

The decision leaves a narrower result than a general approval of Google’s use of publisher content. It dismissed the federal claims for the reasons stated in the opinion. After doing so, the court declined supplemental jurisdiction over the California unjust-enrichment claims. It did not resolve their merits. The opinion states that a final, appealable order accompanies it.[14]

For future litigants, the work begins with the relationship between the challenged conduct and the market allegedly restrained. A reciprocal-dealing theory needs facts supporting the claimed arrangement. A tying theory needs a defensible account of separate demand. A monopoly-maintenance theory needs facts connecting the plaintiff’s injury to exclusion in the relevant market. These requirements cannot be satisfied by using dependence as a substitute for analysis.

For courts and lawmakers, however, dependence is a reason to look more closely. A platform’s authority to determine the conditions of distribution can shape whether its suppliers have any practical opportunity to negotiate over new uses of their work. Rules addressing that problem could protect a publisher’s ability to withhold specified AI uses without sacrificing ordinary search eligibility. Such a proposal would need clear technical boundaries and a remedy available to affected publishers. Otherwise, its value would depend on the willingness of a public enforcer to act in each dispute.

My concern is the durability of that remedy. Publishers should be able to invoke a workable legal right when a dominant intermediary uses control of distribution to impose terms that unlawfully impair competition. The September 30 decision shows why that right requires careful pleading under existing law. It also shows why preserving independent production of information may require a clearer legal account of the power that connects distribution, content acquisition, and competition.

Notes

---

[1] Chegg, Inc. v. Google LLC and Penske Media Corp. v. Google LLC, Nos. 25-cv-00543, 25-cv-03192, slip op. at 1–2, 17–41 (D.D.C. Sept. 30, 2026) (granting motions to dismiss the amended complaints).

[2] Chegg and Penske Media, slip op. at 3–7 (describing the publishers’ businesses and their alleged exchange of indexable content for search referrals).

[3] Chegg and Penske Media, slip op. at 2, 7–12, 16 (recounting allegations about snippets, AI products, and lost referrals while applying the pleading standard).

[4] Chegg and Penske Media, slip op. at 17–23 (rejecting the pleaded reciprocal-dealing theory for lack of a plausible agreement).

[5] Chegg and Penske Media, slip op. at 23–25 (recognizing that Section 2 reaches independent action while requiring a dealing relationship for the specific reciprocal-dealing claim alleged).

[6] Chegg and Penske Media, slip op. at 25–28 (rejecting Penske’s tying theory after assuming without deciding that the per se test applied).

[7] Johnson v. Commission on Presidential Debates, 869 F.3d 976, 980–82 (D.C. Cir. 2017) (distinguishing constitutional standing from claim-specific antitrust standing). Chegg and Penske Media, slip op. at 29–34 (dismissing monopoly-maintenance claims for insufficient allegations of antitrust standing).

[8] Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 487–89 (1977) (requiring an injury that reflects the anticompetitive feature of the challenged conduct and rejecting recovery for profits lost through preserved competition).

[9] Chegg and Penske Media, slip op. at 29–31 (rejecting the publishers’ supplier and purchaser theories of participation in general search).

[10] Blue Shield of Virginia v. McCready, 457 U.S. 465, 472–85 (1982) (recognizing a subscriber’s right to sue where the denial of reimbursement was integral to the alleged restraint and presented no danger of duplicate recovery).

[11] Chegg and Penske Media, slip op. at 31–34 (finding insufficient allegations that the publishers’ injuries were indispensable to Google’s acquisition or maintenance of search dominance).

[12] Chegg and Penske Media, slip op. at 35–38 (expressing doubt about standalone nonmonopolistic leveraging while dismissing the theories on market-definition and related pleading grounds).

[13] Chegg and Penske Media, slip op. at 38–40 (rejecting the proposed educational-publishing market and finding insufficient allegations of Google’s participation and dangerous probability of monopolization).

[14] Chegg and Penske Media, slip op. at 40–41 (declining supplemental jurisdiction over the unjust-enrichment claims and identifying an accompanying final, appealable order).