Federal Judge Dismisses Publisher Antitrust Cases Against Google AI Overviews
On October 1, 2026, US District Judge Amit Mehta dismissed a pair of antitrust lawsuits that had been closely watched by anyone who earns a living from search referral traffic. The cases, filed in 2025 by education platform Chegg and media conglomerate Penske Media, alleged that Google's AI Overviews and related generative products violated antitrust law by harvesting publisher content to answer user queries directly, suppressing clicks back to the original sources. Mehta disagreed. His ruling concluded that Google's conduct, as described in the complaints, does not amount to an antitrust violation.
The decision lands with particular weight because of who issued it. Mehta also presided over the Department of Justice's monopoly case against Google's search business, decided in 2024, in which the court found that Google had illegally maintained its search monopoly. A judge already willing to rule against Google on search competition has now found that the company's AI answer products, at least on the theories publishers advanced, do not cross the same legal line. That contrast is the story here.
For publishers who spent the past two years watching referral traffic erode as AI Overviews expanded across the results page, the dismissal removes one of the few legal levers they had pulled. It does not, however, resolve the underlying economic conflict. If anything, it sharpens the question of what recourse remains.
Why the Legal Arguments Failed to Persuade the Court
The plaintiffs built their cases on two distinct claims. Chegg argued that Google illegally scraped its educational content, feeding that material into Gemini models that could then recreate the substance of Chegg's offerings and absorb the traffic that would otherwise have flowed to the site. Penske Media, whose portfolio includes Rolling Stone and Variety, framed the harm in structural terms: sites indexed for organic search were simultaneously having their content harvested for AI answers, with no mechanism to opt out.
Read next Laika's Wildwood: Stop-Motion Fantasy at TIFF 2026The opt-out grievance is where the ruling turned. Mehta's reasoning, as reported, centered on the gap between what publishers expected and what Google actually agreed to. The plaintiffs pleaded an "expectation" that Google would continue sending them search traffic. The court found that expectation legally irrelevant — an expectation is not a contract, and it is not the kind of agreement antitrust doctrine requires.
That distinction matters more than it may first appear. Antitrust claims typically require proof of an unlawful agreement, exclusionary conduct that harms competition in a defined market, or both. A publisher's belief that a decades-old traffic relationship would persist does not establish that Google promised anything enforceable. Media law and antitrust attorneys have made this point repeatedly in commentary on the AI-scraping litigation wave: US antitrust law protects competition, not individual competitors' revenue models. A business whose distribution depends entirely on another company's platform bears the risk when that platform changes. Courts have been consistent on this for years, and Mehta's ruling extends that logic into the generative AI era.
The scraping claim faced a similarly steep climb. Chegg's argument that content ingestion into Gemini reduced its traffic describes a harm, but antitrust doctrine asks whether the conduct forecloses rivals from competing — not whether it inconveniences an upstream supplier. Google's implicit bargain with the open web, however one-sided it may feel to publishers, was never codified into the terms the plaintiffs needed.
What This Ruling Means for Publishers Losing Traffic to AI Search
The traffic decline publishers describe is not speculative. Through 2024 and 2025, multiple independent measurement firms documented meaningful drops in referral volume for sites whose content was eligible to appear in AI-generated answers. Similarweb and Semrush analyses of search referral patterns, along with Cloudflare Radar data on crawler behavior, have tracked the shift repeatedly. Publisher disclosures during earnings calls across 2024 and 2025 described double-digit declines in search-sourced sessions for some content categories, with reference, recipe, education, and how-to verticals hit hardest. Those numbers come with methodology caveats, but the direction has been consistent across sources.
Chegg's situation illustrates the exposure. An education platform whose core product answers questions that a chatbot can now answer directly occupies precisely the territory AI Overviews compress. Penske's titles face a different version of the same problem: entertainment and culture coverage that AI summaries can digest without sending a reader to the review.
The dismissal means publishers cannot expect antitrust courts to restore that traffic. The legal theory that Google's market power obligates it to keep sending clicks failed. Whatever remedies emerge will come from legislation, regulation, negotiated licensing, or publishers changing their own distribution strategies — not from this litigation track.
The Opt-Out Problem: Can Publishers Protect Their Content?
Penske's opt-out argument identified the sharpest asymmetry in the current arrangement: a site can be indexed for traditional search while its content is simultaneously pulled into AI answers, and it has no meaningful switch to separate the two. The court found no legal violation in that structure. The practical question is what publishers can do instead.
Technical options exist but carry tradeoffs. Robots.txt directives and crawler-level controls can block AI ingestion, though blocking Google's AI crawlers risks collateral damage to search visibility, since the same infrastructure serves both. Some publishers have pursued licensing deals with AI developers, trading content access for compensation. Others have leaned into paywalls, newsletters, and direct audience relationships that do not depend on search referral at all. Each approach has costs, and none restores the volume that search once delivered.
The opt-out debate now shifts to legislatures and regulators. If courts will not recognize an implicit right to control AI ingestion of indexed content, the only path to a formal opt-out runs through statute.
What Comes Next: Legal, Legislative, and Business Options
The dismissal is not necessarily the final word. Plaintiffs can appeal, and appellate courts may read the agreement question differently. But appeals are slow, and the underlying doctrinal problem — expectation versus enforceable commitment — would follow the case upward.
Legislative activity is the more likely near-term pressure point. Proposals touching on AI content ingestion, transparency, and compensation have circulated at the federal and state levels, and this ruling gives proponents a concrete example of why statutory fixes are needed. Whether that momentum converts into law is a separate question, and past attempts have stalled amid competing industry interests.
The business track may move fastest. Licensing negotiations between publishers and AI developers are already underway in various forms, and the dismissal removes the litigation threat that gave some publishers leverage at the table. That cuts both ways: without a courtroom alternative, deals may be struck on terms less favorable to content owners.
The Bigger Picture: AI Search and the Future of the Open Web
Step back, and the ruling clarifies the terms of a much larger contest. The open web's economic model rested on an implicit exchange: publishers made content freely crawlable, and search engines sent traffic in return. AI answers weaken the second half of that bargain while the first half remains fully in force.
Mehta's ruling says that imbalance, as currently structured, is not an antitrust problem. Publishers who built their businesses on search referral now face a market where the referral is no longer guaranteed and the legal system offers no remedy for its disappearance. The next phase will be fought over legislation, licensing terms, crawler standards, and audience strategy — quieter battlefields than a federal courtroom, but the ones that will actually determine whether independent publishing survives the transition to AI-mediated search.
Source: Ars Technica - All content
