Publishers have a real grievance, and Judge Amit Mehta's ruling this week did not dispute it. It did hold that the grievance is not an antitrust claim as pleaded. Mehta dismissed the amended complaints that Penske Media and Chegg filed against Google over AI Overviews. Coverage appeared on October 1–2, 2026. The ruling is a useful check on a growing instinct to treat every disruptive product launch as a competition violation.
The strongest case for the publishers
The publishers' theory deserves a fair statement. Google is a court-adjudicated search monopolist. The same judge found as much in the government's search case, and the Justice Department's case page lists a Final Judgment and Memorandum Opinion dated December 5, 2025. Penske's September 2025 complaint in the D.C. federal court pleaded six counts, including reciprocal dealing and monopoly leveraging under the Sherman Act. It alleged that Google conditions search referral traffic on publishers also supplying content for snippets, Gemini training and AI Overviews. In this telling, a publisher that wants to be found cannot say no, and the summary then answers the query without sending the reader onward. That is a coherent story about market power, and Mehta himself acknowledged what it describes. According to PPC Land's account, he wrote of the "knock-on consequences to journalists, educators, and other online creators whose content Google takes and repurposes without compensation."
Why the claims failed
The legal failure was specific. Per reporting on the order, the court rejected five categories of claims:
- Reciprocal dealing: the publishers pleaded no actual agreement. Mehta wrote that "an expectation is not an agreement." Publishers also "failed to plead any actual agreement whereby Defendants promised to 'sell' Plaintiffs any specific amount of traffic," per PPC Land.
- Tying: the plaintiffs could not show that search and AI Overviews are separate products with distinct consumer demand.
- Monopoly maintenance: the publishers lacked antitrust standing in the general search services market.
- Attempted monopolization and leveraging: no plausible markets were defined and no dangerous probability of success was shown.
- California unjust enrichment: the court declined supplemental jurisdiction once the federal claims fell.
The tying holding matters most for the technology sector. Google argued in its January 12, 2026 dismissal motion that AI Overviews are not a separate product. It described them as another way of presenting search results, comparable to featured snippets. It also argued that search traffic results from user decisions, not from Google selling a product. The court's reasoning tracks that logic. If every new interface element on a platform were a separately tied product, product design would be an antitrust question, and incumbents and challengers alike would have to litigate each layout change.
Antitrust protects competition, not expected revenue
The Federal Trade Commission's guide to the antitrust laws says the laws exist to "protect the process of competition for the benefit of consumers." The Sherman Act's monopolization provisions reach "any monopolization, attempted monopolization, or conspiracy or combination to monopolize." None of that language gives a business a claim to a particular volume of referral traffic. Search traffic was never a purchased good. It was a by-product of how a general search engine ranks and displays results, and Mehta's phrase captures exactly that.
The pro-innovation reading is straightforward. A rule that a platform must preserve the traffic economics of a prior product format would freeze that format in place. Search engines have always summarized and re-presented information, through snippets, knowledge panels and answer boxes. Treating the AI version as categorically different, without a showing of foreclosed competition, would give legacy publishers a veto over how new products evolve. That would hurt the AI entrants and open-web tools that depend on the same freedom to build.
What the ruling does not settle
The decision is narrower than some headlines suggest. The Penske dismissal is reported as without prejudice and as final and appealable, so refiling or an appeal remains possible. The reports reviewed did not specify the prejudice status of Chegg's dismissal. Nothing in the ruling declares AI Overviews pro-competitive, and nothing resolves copyright questions, which sit in different statutes and different courts. Nor does it say that Google's conduct toward publishers is fair. The judge's own acknowledgment of the harm suggests the opposite.
Google's own filing points to where the real answer lies. It noted "a booming market for AI content licensing," in which publishers strike multiple deals. That market, not an antitrust decree, is the more proportionate route to compensation. It rewards content that AI developers actually need and leaves price discovery to the parties. If there is a persistent gap, such as publishers with no practical way to opt out of AI use without leaving search, the right tools are narrower and more evidence-driven. Examples include transparent opt-out controls, clearer crawler separation, and disclosure of how referral traffic changes over time. Those fixes are better pursued through legislation or targeted enforcement than by stretching the Sherman Act to cover an expectation.
The takeaway for policymakers
The dismissal is a reminder that antitrust is an instrument with defined elements: agreement, product markets, standing, and harm to competition rather than to individual competitors. When a plaintiff cannot satisfy them, the answer is not to loosen the elements. It is to ask whether the underlying problem needs a different tool. Congress can legislate on content licensing, transparency or opt-out rights if it concludes that the market is failing. Until it does, courts applying existing law will keep finding that disruption alone is not a violation. For a technology sector that depends on predictable rules, that restraint is a feature.