Google Wins Dismissal Of Penske Media, Chegg AI Lawsuits

The lawsuits

Penske Media, whose properties include Rolling Stone, Variety, The Hollywood Reporter, Billboard and Deadline, sued Google and its parent company Alphabet in September 2025. Chegg, which provides online educational resources and operates a large database of question-and-answer material, filed its lawsuit earlier that year.

The cases developed around a common argument: Google had historically operated as an intermediary between internet users and publishers, crawling websites and displaying links that directed users to the publishers’ pages. The publishers, in turn, supplied Google with the material necessary to create its search index.

According to the plaintiffs, that relationship changed as Google increasingly incorporated content directly into its search results, first through features such as featured snippets and later through generative-AI products such as AI Overviews.

The plaintiffs alleged that Google was effectively requiring them to make their content available for AI-related uses as a condition of continuing to receive the search exposure on which their businesses depend.

Penske’s complaint characterized the traditional relationship as an exchange in which publishers allowed Google to crawl their material while Google sent users back to publishers’ websites. Penske alleged that Google’s AI products altered that arrangement by using publishers’ material to generate answers directly on Google’s search-results page, reducing the incentive for users to click through to the original sources.

Chegg made a similar argument from the education sector. Its complaint described its proprietary database of roughly 135 million question-and-answer solutions as a core asset and alleged that Google could use educational publishers’ material to generate competing answers through its AI systems.

The court’s opinion says Chegg had invested “hundreds of millions of dollars” in the human and technological resources used to create and maintain its content, according to the allegations in its amended complaint.

What the companies alleged

Both plaintiffs pursued theories under Sections 1 and 2 of the Sherman Act.

At the center of their cases was a theory of “reciprocal dealing.” They argued that Google was conditioning the provision of search referral traffic on publishers supplying content for uses beyond traditional search indexing.

The plaintiffs identified three categories of content they said Google was obtaining without payment: material that could be republished in search results, material used to train Google’s large language models, and material used to generate or ground AI responses.

They also identified a market for search referral traffic and argued that Google was using its dominance in general search to obtain those inputs at no cost.

Penske separately alleged that Google was unlawfully tying its general search product to AI Overviews. Under that theory, general search was the tying product and AI Overviews the tied product.

Both companies also alleged that Google was maintaining its search monopoly through the challenged practices and attempting to extend its dominance into other markets.

Penske claimed the relevant downstream market was online publishing. Chegg defined a separate market for online educational publishing.

The companies also brought California unjust-enrichment claims.

The judge’s central problem with the cases

Mehta’s first major conclusion was that the plaintiffs had not plausibly alleged that Google and the publishers had an agreement of the kind required for their reciprocal-dealing theories.

The companies argued that an agreement could be inferred from the long-standing relationship between Google and publishers: publishers allowed Google to crawl their material, and Google provided search traffic in return.

But the judge said an expectation of receiving traffic is not the same thing as an agreement.

“An expectation is not an agreement,” Mehta wrote, describing the plaintiffs’ theory as simply the way a general search engine operates.

The court also rejected the argument that the parties’ “historical course of dealing” established an implied agreement. Mehta said such a course of dealing would still have to plausibly establish essential terms of an agreement, such as price or quantity.

The complaints, he concluded, did not allege that Google and the publishers had reached a meeting of the minds over the exchange.

The judge also pointed to the enormous implications of accepting the plaintiffs’ theory. If allowing Google to crawl a website while expecting search traffic constituted a contractual relationship, he wrote, the theory could imply that Google has agreements with billions of websites it indexes.

The court found that implausible. Publishers, it said, keep their websites open to Google’s crawlers because doing so serves their own interest in receiving search traffic, not because they entered into a commercial bargain with Google.

AI Overviews did not establish an unlawful tie, the court said

Penske’s tying claim also failed.

The company argued that Google was using its monopoly in general search to force users to accept AI Overviews. But Mehta concluded that Penske had not plausibly alleged that Google Search and AI Overviews are separate products under antitrust law.

The court treated AI Overviews as part of an integrated search experience rather than a separate product that consumers are compelled to purchase.

Mehta noted that Google Search can produce many types of results, including ordinary links, advertisements and other specialized results. If AI Overviews were treated as a separate product merely because they appear on the search-results page, the same logic could potentially make each type of search result a separate product.

The judge also found an internal tension in Penske’s argument. The company said Google Search is intended to send users to other websites while AI Overviews are designed to keep users on Google’s results page. But, Mehta observed, Penske’s own allegations said users who receive satisfactory answers from AI Overviews have less reason to visit outside websites.

That, the judge said, suggested that both products serve the same basic consumer demand: providing answers to search queries.

The tying claim was therefore dismissed.

The standing problem

Another major issue was antitrust standing.

The plaintiffs argued that they should be permitted to sue because they participate in the general search ecosystem by supplying the content Google indexes and receiving search traffic from Google.

Mehta rejected that theory.

Under antitrust law, a private plaintiff must generally show that it suffered the type of injury the antitrust laws are designed to prevent and that the injury occurred in the market where competition is allegedly being restrained.

The judge concluded that Penske and Chegg had not made that showing with respect to Google’s general search market.

Their alleged losses — including lost subscription revenue and inadequate compensation for the use of their content to train or ground AI systems — occurred in the publishing markets, Mehta said, rather than in the general search market itself.

The plaintiffs also argued that their injuries were sufficiently connected to harm suffered by Google’s AI and search competitors to confer standing.

The court rejected that argument as well.

Mehta said the complaints did not plausibly establish that obtaining publisher content for free was an indispensable part of Google’s ability to maintain its monopoly in general search. Google had already established its dominant position in search before the emergence of generative AI, the judge noted.

The court cited earlier findings that Google handled approximately 80% of general search queries by 2009 and 89.2% by 2020. AI Overviews, meanwhile, became broadly available to U.S. users only in 2024.

The plaintiffs did not allege that Google had uniquely obtained publisher content for free while its search competitors paid for equivalent material, or that the alleged cost difference was sufficiently large to produce anticompetitive effects in general search, according to the ruling.

The proposed publishing markets also failed

The court separately found problems with the plaintiffs’ definitions of the markets in which Google allegedly was attempting to gain or maintain monopoly power.

Penske defined “Online Publishing” broadly enough to encompass essentially all digital, text-based content published online.

Mehta found that definition implausibly broad. A blog post, legal brief, fictional story, news article and Chegg’s question-and-answer database, for example, could not simply be presumed to be interchangeable products in one market, he wrote.

Penske also had not alleged Google’s market share in that proposed publishing market or provided facts establishing that Google faced a dangerous probability of monopolizing it.

Chegg’s proposed online educational publishing market was more narrowly defined, but the court still found it too vague.

Chegg had argued that educational publishing was distinguished by attributes including curation, verification, authority and pedagogical focus. Mehta said the company had not provided enough factual specificity to determine how those characteristics separated educational content from other nonfiction material.

The court also questioned whether Google itself was a participant in the market as Chegg had defined it.

Because the plaintiffs failed to establish plausible relevant markets, their attempted-monopolization and monopoly-leveraging claims could not survive.

The court acknowledged the broader problem

Perhaps the most consequential passage in the ruling came as Mehta addressed the larger concerns underlying the litigation.

The judge said he did not “treat Plaintiffs’ alleged harms lightly” and was “not unsympathetic” to the position of publishers whose content Google takes and repurposes without compensation.

He specifically referenced the potential consequences for journalists, educators and other online creators.

But Mehta said antitrust law has limits.

The antitrust statutes, he wrote, are not a substitute for legislative action addressing “economic dislocation caused by new innovation.” If the litigation exposed deficiencies in the reach of existing antitrust law, he said, those were questions for Congress or regulators to consider.

“The court is bound to apply the law only as it is written,” Mehta wrote, “not how the Court or any party believes it ought to be.”

That distinction is central to the ruling. The decision does not resolve the broader policy debate over whether publishers should be compensated when their work is used to develop or power generative-AI systems. Instead, it holds that the particular antitrust theories advanced by Penske and Chegg were not adequately pleaded under existing law.

Google wins both cases

After rejecting the federal antitrust claims, Mehta also declined to exercise supplemental jurisdiction over the plaintiffs’ California unjust-enrichment claims.

The court therefore granted Google’s motion to dismiss Chegg’s amended complaint and its motion to dismiss the amended complaint filed by Penske Media and its affiliated publishers. A final, appealable order accompanied the opinion.

The Penske case involved the company and a group of affiliated publishers including Billboard Media, Deadline Hollywood, Fairchild Publishing, Gold Derby Media, The Hollywood Reporter, IndieWire, Penske’s Rolling Stone operation, SheMedia and Variety Media.

The lawsuits had placed a major media company and a major education platform on the same side of a growing fight over how the economics of the internet should work as search engines become answer engines.

Penske had argued that publishers face a fundamental choice: allow Google to crawl their material and risk having that material repurposed by AI systems that can answer users’ questions without sending them to the original source, or restrict Google’s access and risk losing the search visibility on which their businesses depend.

Google disputed that legal characterization and argued that its AI features are part of its search product. In seeking dismissal, Google also argued that publishers are not guaranteed referral traffic and can restrict Google’s crawling of their websites.

For now, Mehta’s ruling leaves that larger dispute outside the scope of these two antitrust cases.

The decision could nevertheless have implications beyond Penske and Chegg, as publishers, technology companies and policymakers continue to grapple with questions over AI training, search traffic, copyright, compensation and the changing economics of the open web.

For Penske and Chegg, however, the immediate result is clear: their federal antitrust challenges to Google’s AI search practices have been dismissed, with the court concluding that the complaints did not satisfy the requirements of existing antitrust law.

Follow me at TooMuchTV.com and subscribe to the award-winning Too Much TV newsletter.