Reality bites. And the great debate over whether publishers will opt out of Google Search is not reality. 

In recent months, I’ve watched everyone from Adweek and NiemanLab to The New York Times and The Wall Street Journal posit this question. This week, Digiday devoted an hour-long podcast to pros and cons. 

Let me be the first to say it bluntly and publicly: It ain’t happening. 

Diversifying discovery and business away from Google? Absolutely. Every smart publisher is working to do it. But opting out of Google Search is an entirely different proposition.  Breaking away is existential. And you’re delusional to suggest any investor is going to fund a martyr to do it. 

Let’s review the data. 

Just in case it has slipped anyone’s mind, Google has been found to have illegally monopolized the key advertising pipes to monetize the open web along with Google’s app store. And the seed to both of those adjudicated monopolies began with its illegal monopoly over search of the open web. 

Judge Amit Mehta found Google illegally maintained its search monopoly, with 89% of general search queries and a mobile share pushing 95%. Google’s capture of queries and clicks at a scale more than 10x (an order of magnitude) larger than its nearest competitor allows it to improve its responses, particularly across ‘tail queries,’ in a manner no one else can. And given its apparent mastery of monopolies, Google locked in the search market share it already owned through its Chrome browser and Android operating-system distribution while sending tens of billions in cash per year to the one company that could potentially displace it, Apple, to secure default search distribution across Apple’s other half of the mobile market. 

So, Judge Mehta got the liability decision right. Very much so. 

The illusion of choice 

Here’s something else Mehta got right: to use his own words in his remedies opinion (pg 204), “Publishers cannot, however, opt out of Google’s use of their content to fine-tune Google’s Search models or for display in AI Overviews.” Mehta described publishers as caught “between a rock and a hard place” as AI Overviews are bundled with search. Mehta described opting out of Google crawling altogether “not a tenable choice.” Exactly. So, let’s stop debating.  

Other AI companies face a different reality. Publishers can block them, negotiate with them or sue them. Increasingly, AI companies are licensing premium content. There is clearly a market for those rights. 

Google has something they don’t: the dominant business for crawling the open web. It has piggybacked AI Overviews onto Search, substituting AI-generated responses for publisher visits while giving publishers no meaningful ability to withhold their content without jeopardizing search traffic. 

If you’re reading this and are a member of DCN, you’ve seen our data at our last AI member event and know we’ve continued to privately study the effects. Yes, Google traffic has been dropping for two years. But it’s still the largest source of traffic outside direct. Organic search – again, 90%+ coming from Google – still represents roughly 23% of inbound traffic across a representative set of DCN members, down from 25% a year earlier.  

Twenty-three percent of traffic. Show me the CEO or CFO who will voluntarily turn that off. 

Diversification in discovery isn’t coming by strategic choice as much as some press-friendly publishers would like you to believe. It’s coming because Google is systematically substituting for the open web, hijacking the profits while sitting atop its adjudicated illegal monopoly over its most important discovery mechanism.  

The Penske lawsuit and the antitrust question 

So, what do we do about it? 

That question brought me to Judge Mehta’s all too familiar Courtroom 10 on this Tuesday afternoon late in August for oral arguments in Penske Media v. Google. Google questions its novel approach to the law, but the business case is very clear. Anyone in our industry would be smart to give it a read. Penske argues Google has broken what Google itself has called the “fundamental fair exchange between Google and the web.” Publishers let Google crawl their content, and Google sent users back to them.  

Publishers today still need the search traffic, but Google has conditioned access to it on the ability to use the publishers’ content for new purposes, including AI training and AI Overviews, without paying for those additional rights. Penske argues that Google getting those inputs without paying not only harms publishers but raises costs and barriers to entry for Google’s AI competitors. Exactly.  

Google’s lawyers had a very different almost humorous (if it weren’t insulting) description of this history in court on Tuesday. They characterized Penske’s description of the original traffic-for-crawling bargain as a “vague historical course of dealing.” AI Overviews, they repeatedly argued, are simply a “product improvement.” And the idea that this could all eventually be bad for publishers? They called that “way too speculative.” 

That’s when things got deeply interesting and had me leaning in: Mehta acknowledged that Google’s approach does not allow publishers to control how their content is used even saying he has “sympathy” for the problem. Then he asked why Google wouldn’t be subject to antitrust scrutiny for this. [It] “all seems really unfair,” Mehta said. 

At another point, in response to Google’s generalizing of its AI Overviews as product improvement, Mehta said it has been done “on the backs of the publishers.” Exactly. And importantly, he observed that these product improvements “are not immune from antitrust scrutiny.” While calling it beyond his pay grade, he even suggested the situation seemed in some respects like an “essential facility.”  

That gets us much closer to the real issue. Penske, and the other publisher plaintiffs, aren’t merely complaining that Google is stealing traffic as Google proxies would like everyone to believe. Penske argues Google is using its search monopoly to distort the emerging market for generative AI inputs. 

Mehta himself observed Tuesday that a market is clearly forming for GenAI inputs. Penske’s argument is that publishers would license content to Google’s AI products too – except for the glaring issue of Google’s ability to condition access to search on its ability to use their content for other purposes. As Penske’s lawyer put it Tuesday, publishers “by and large can’t say no to this.”  

The house always wins 

Google’s structural advantage isn’t simply its 90%+ share of search. It uses that monopoly to acquire the very inputs required to train and improve the AI products now supposedly threatening its monopoly, all without paying the market price its nascent competitors have to pay. This means that Google still takes the content for free, but increasingly uses it for additional purposes including AI training, grounding and republishing that reduces the output publishers receive in the form of traffic. 

That’s quite a business model. Now let’s return to the question everyone seems to be asking:  

Will publishers opt out of Google? 

No. That’s the headline. 

For two decades, Google has argued publishers voluntarily give it access to scrape their content because Google sends them traffic. Competition on the web is merely a click away. Now that Google is increasingly keeping that audience for itself, its only defense depends on the notion that publishers remain free to leave its search experience and this recently narrative that they just might.  

Judge Mehta has already recognized that this isn’t a reality. The important question isn’t whether publishers will walk away from Google. It’s whether an adjudicated monopolist gets to use the fact that publishers can’t walk away to dictate the terms of the next internet. 

If the answer is yes, stick a fork in the open web.