[The Engines]
How Antitrust Scrutiny Could Impact Google's AI Content Practices
A federal judge just told Google its AI content deal with publishers looks 'really unfair.' Here is what changed, who it affects, and how to respond before the case is decided.
Explore this article with AI
Open a source-aware analysis with this article as the primary source.The short answer
On August 26, 2026, Judge Amit Mehta called Google's AI content deal with publishers 'really unfair' during a motion-to-dismiss hearing in Penske Media Corporation v. Google LLC. No ruling has come down yet, but the case tests whether Google's crawl-for-traffic arrangement with publishers is itself anticompetitive. If your content depends on being indexed and cited, the outcome will shape how much control you actually have over that citation.
What changed in Google's AI content deal with publishers?
On August 26, 2026, federal judge Amit Mehta heard arguments on Google's motion to dismiss Penske Media Corporation v. Google LLC in the U.S. District Court for the District of Columbia, and he did not sound convinced by Google's defense. Mehta called the arrangement between Google and publishers 'really unfair' and said Google is building AI Overviews 'on the backs of the publishers,' according to Search Engine Journal's account of the hearing.
Penske Media, which owns more than 25 print and digital properties including Rolling Stone and Billboard, filed the suit in September 2025. Its central claim: Google conditions search indexing on publishers allowing AI-generated summaries of their content, and those summaries cut the referral traffic that funds newsrooms. Google's motion to dismiss argues the opposite framing. AI Overviews are a 'product improvement' to search, not a separate product, and the long-standing deal that lets Google crawl a site in exchange for traffic is, in Google's own words, a 'vague historical course of dealing,' not an enforceable contract publishers can renegotiate.
Google also told the court that Penske's preferred outcome would leave Search 'frozen in time,' per PYMNTS' coverage of the motion filings. Mehta was not fully persuaded. He noted that product improvements 'are not immune from antitrust scrutiny,' and pressed on whether Google's search monopoly, the same monopoly he found illegal in the Department of Justice's 2025 case, lets it take publisher content for AI training and AI answers at no cost while OpenAI and Perplexity negotiate and pay for comparable material.
No ruling has come down as of this writing. Chegg has filed a related case making a similar traffic-loss argument against Google's AI Overviews, so Penske is not an isolated plaintiff. This is a pattern of litigation now, not a single lawsuit.
Who does this antitrust fight affect?
Two groups feel this fight first, and a third feels it indirectly.
B2B SaaS and tech companies competing for the 'best [category] tool' prompt depend on Google indexing their content before ChatGPT, Perplexity or Gemini ever see it. Most AI engines still lean on search infrastructure to find and rank sources. If Google can rewrite crawl terms unilaterally to favor its own AI Overviews, the on-ramp every other engine relies on narrows too, which is exactly the anticompetitive harm Penske is arguing in court.
Publishers and content-heavy businesses face the sharper version of the same problem, and the numbers back up their complaint. A Shero Commerce study of 1,851 citations across Google AI Mode, ChatGPT and Perplexity found that brand-owned pages captured only 2.8% of citations. Even when a brand was recommended by name, its own page was the cited source just 31% of the time, according to Search Engine Journal's write-up of the research. Most of the time, a review site, forum or aggregator earns the citation instead of the brand actually being discussed.
Local and multi-location service businesses sit a step removed from the lawsuit, but they run on the same indexing pipeline every time someone asks an engine for the best electrician, dentist or accountant nearby. A ruling that reshapes how Google licenses or restricts crawling would ripple into every '[service] in [city]' answer, not just publisher news content.
| Dimension | Before Aug 26, 2026 | After Aug 26, 2026 | What to do now |
|---|---|---|---|
| Legal exposure | Google's product improvement defense was untested in court. | Judge Mehta called the arrangement really unfair and said product improvements are not immune from antitrust scrutiny. | Do not assume Google's current crawl-for-citation terms are permanent policy. |
| Opt-out framing | Google argued publishers can simply opt out of indexing if they dislike the terms. | Mehta noted opting out means disappearing from Google search entirely, undercutting the choice defense. | Build citation share on engines that do not depend on Google's index terms. |
| Licensing gap | OpenAI and Perplexity pay publishers for content; Google largely has not. | That gap became a central point in Penske's antitrust argument during the hearing. | Treat any citation you earn as owned proof, not something rented from one company's index. |
| Legal precedent | No prior ruling connected Google's search monopoly to its AI content practices. | The same judge who found Google's search monopoly illegal in 2025 is now applying that finding to AI Overviews. | Watch the ruling on the motion to dismiss, but publish as if the case will not save you either way. |
Why is Judge Mehta skeptical of Google's defense?
Mehta's skepticism centers on control, not on Google's intent.
Google's fallback argument is that publishers can always opt out of indexing if they dislike the terms. Mehta pushed back on that framing directly: opting out means disappearing from Google search entirely, which is not a real choice for a business that depends on search traffic to survive. That distinction, between a technical option and a viable one, is the crux of Penske's antitrust theory.
The framing carries extra weight because Mehta already ruled, in the Department of Justice's separate 2025 case, that Google holds an illegal monopoly in general search. Applying that same lens to AI Overviews means asking whether Google can use monopoly-scale index access to source free inputs for its AI products while competitors pay for the same content. Penske's lawyers drew exactly that comparison during the hearing, pointing to OpenAI's and Perplexity's licensing deals with publishers as the market rate Google avoids paying.
None of this settles the case. A skeptical question from the bench is not a verdict, and Mehta has not ruled on the motion to dismiss. What the hearing confirms is that the 'it's just a product improvement' defense is getting real scrutiny in federal court, in a case publishers across the industry are watching alongside Chegg's parallel suit.
How should you respond to this antitrust uncertainty?
Do not wait for a ruling before changing how you show up in AI answers. This case will take months to resolve, longer if Google appeals, and the underlying dynamic, being crawled and quoted with little control over the outcome, predates the lawsuit and will outlast it.
Four moves hold up regardless of how the case ends.
The honest version of this advice is not that any business can control Google's legal exposure. Nobody can, including Google's own lawyers, who are still arguing the point in court. What a business can control is whether it has enough citable material, spread across enough engines, that one company's litigation is not the difference between being found and being invisible.
- Publish content specific enough that an engine has to cite you directly, not a summarizer. Generic brand pages lose that fight to review sites and forum threads most of the time, per Shero Commerce's citation study as reported by Search Engine Journal.
- Attach a real, dated number to your claims. Original data and comparisons are what a model actually quotes, not adjectives.
- Track citation share across ChatGPT, Perplexity, Gemini and AI Overviews separately. A change to Google's crawl policy would not touch how Perplexity or ChatGPT choose what to cite.
- Treat this lawsuit as a live signal, not background noise. If the court forces Google to change licensing terms, businesses already publishing at scale and already tracking citations will adapt fastest.
Does this signal a bigger shift in AI licensing?
The Penske case exposes a wider gap in the AI content economy: some engines pay, and Google, until now, largely has not. OpenAI and Perplexity have both signed content licensing deals with publishers, a strategy that doubles as antitrust insurance since it establishes a market rate for content and is harder to describe as coercive indexing. Google, built on a search monopoly that predates generative AI, has leaned on crawl access instead.
That gap does not mean Google alone determines whether your content gets cited. Citation Share, the percentage of relevant AI answers that cite a given business, gets measured across ChatGPT, Perplexity, Gemini and Google AI Overviews, not one engine in isolation. A court ruling against Google would change the rules for one engine among several a buyer actually consults before deciding.
The bigger lesson from the hearing is structural, not legal: coverage and freshness at scale are what get a source cited, whether that source pays for content, crawls it under a legacy deal, or licenses it outright. Businesses that publish across every major engine are not exposed to a single company's court case the way a Google-only strategy is.
What happens next in Penske Media v. Google?
Judge Mehta has not ruled on Google's motion to dismiss as of this writing. If he denies it, Penske's antitrust claims move into discovery, and Google's crawl-for-citation arrangement with publishers faces scrutiny it has never had to survive before. If he grants it, Google's product-improvement defense holds, at least in this court, and publishers lose their most direct legal lever against AI Overviews.
Either outcome covers one case, in one district court, involving one company's practices. It will not settle how AI engines source and cite content everywhere. What it will do is set a precedent other publishers, Chegg included, and possibly regulators can point to the next time an AI company's citation practices look, in a federal judge's own words, really unfair.
Key takeaways
- On August 26, 2026, Judge Amit Mehta called Google's AI content deal with publishers really unfair during a motion-to-dismiss hearing in Penske Media v. Google.
- Google argues AI Overviews are a product improvement; Penske argues indexing is conditioned on allowing AI summaries that cut referral traffic.
- Mehta already ruled Google holds an illegal search monopoly in the DOJ's 2025 case and is now asking whether that monopoly extends to free AI content inputs.
- Chegg has filed a related suit against Google over AI Overviews, so this is a pattern of litigation, not an isolated case.
- Brand-owned pages captured only 2.8% of AI citations in a 1,851-citation study across Google AI Mode, ChatGPT and Perplexity.
- No ruling has come down yet. Build citation share across multiple engines rather than wait for the case to resolve.
Omnicite Editorial. "AI Content Deal: Google's Antitrust Test" The Citation Report, Omnicite. https://omnicite.co/blog/how-antitrust-scrutiny-could-impact-google-s-ai-/
Sources
Source: Search Engine Journal
Judge Amit Mehta called Google's AI content arrangement with publishers 'really unfair' and said it was built 'on the backs of the publishers' at the August 26, 2026 hearing. Search Engine Journal, 2026-08-27
Source: Search Engine Journal (reporting on Shero Commerce research)
Only 2.8% of 1,851 AI citations tracked across Google AI Mode, ChatGPT and Perplexity pointed to a brand's own page, and brands recommended by name were cited via their own page only 31% of the time. Search Engine Journal (reporting on Shero Commerce research), 2026-08-26
Source: PYMNTS
Google told the court AI Overviews are a product improvement and that Penske's position would leave Search frozen in time; Penske alleges indexing is conditioned on allowing AI summaries that cut publisher traffic. PYMNTS, 2026-01-14
Source: Complete AI Training
Overview of the August 26, 2026 antitrust hearing in Penske Media v. Google and its implications for Google's AI content practices. Complete AI Training, 2026-08-30
Frequently asked questions
What is Penske Media v. Google actually about?
Penske Media, publisher of Rolling Stone and Billboard among other titles, sued Google in September 2025 in the U.S. District Court for the District of Columbia. It alleges Google conditions search indexing on publishers allowing AI-generated summaries of their content, and that those summaries cut the referral traffic publishers depend on.
What did Judge Mehta say about Google's AI content deal?
At the August 26, 2026 hearing on Google's motion to dismiss, Mehta called the arrangement 'really unfair' and said Google is building AI Overviews 'on the backs of the publishers.' He also said product improvements are not automatically immune from antitrust scrutiny.
Has Google already been found to hold an illegal monopoly?
Yes. The same judge, Amit Mehta, ruled in the Department of Justice's separate 2025 case that Google holds an illegal monopoly in general search. He is now weighing whether that monopoly extends into how Google sources content for AI Overviews.
Can publishers just opt out of Google's AI Overviews?
Technically yes, by blocking indexing entirely. Mehta pointed out that this is not a real choice, since opting out means disappearing from Google search altogether, not just from AI summaries.
Is Penske the only company suing Google over this?
No. Chegg has filed a related case alleging similar traffic losses tied to Google's AI Overviews, so the legal pressure on Google's AI content practices is coming from more than one plaintiff.
When will the court decide whether the case can proceed?
No ruling had come down as of the reporting on the August 26, 2026 hearing. A denial of Google's motion to dismiss would send the case into discovery; a grant would end it, at least at this court.