# How Do You Kill a Parasite? > Source: > Published: 2026-07-27 12:36:47+00:00 TL;DR — Key Takeaways - Google’s AI Overviews and AI Mode increasingly use publisher content to answer questions directly, keeping users—and advertising value—inside Google. - Individual publishers risk major traffic losses if they block Google alone, making coordinated action essential. - A proposed publisher compact would require separate controls for search and AI, affirmative consent, usage transparency, attribution and compensation. In my last article, [Google Is a Parasite. And Everyone Else Is Paying the Price](https://docs.google.com/document/d/1jJ3q2yytZVcb5Q35SwDMG5Dao3jLBIgGVpnSnbmg9rw/edit), I argued that Google had crossed a line in its relationship with the open web. For years, Google and publishers existed in something resembling a symbiotic relationship. Publishers allowed Google to crawl and index their work. Google organized it and sent people back to the original source. Google built an extraordinarily profitable advertising business. Publishers received the audience, advertising opportunities and customer relationships they needed to finance the creation of more content. AI Overviews and AI Mode have fundamentally rewritten that bargain. Google still consumes the work, but it increasingly uses that work to construct the answer itself. It keeps the user, the advertising opportunity, the data and the customer relationship. The publisher finances the reporting, research and expertise. Google harvests it, packages it and decides how much value—if any—to return. That was the diagnosis. Now comes the harder question: How do you kill a parasite? The blunt answer is that you stop feeding it. The problem is that one publisher cannot safely do that alone. ### The First Signs of a Revolt A recent [Wall Street Journal report](https://www.wsj.com/business/media/google-search-publishers-ai-content-0fb06e41) suggests some publishers have begun seriously considering the previously unthinkable: Cutting Google off. USA Today, Politico, Reuters, The Economist and People Inc. are among the publishers reconsidering how—or even whether—they should continue working with Google. Reddit is reassessing a reported $60 million-a-year content arrangement with Google as that deal approaches renewal. People Inc. CEO Neil Vogel told the Journal that turning Google off is “100% on the table.” USA Today Co. CEO Mike Reed was even more direct: “It’s time to take a stand and say enough is enough.” That reporting did not remain confined to the media-business pages. Barry Schwartz picked it up at [Search Engine Roundtable](https://www.seroundtable.com/publishers-blocking-google-search-41735.html), a publication followed closely by the search and SEO communities. That matters. This is no longer just a group of disgruntled publishers complaining that Google has changed the rules. It is becoming a fundamental question about whether Google Search can continue functioning if the sources it depends upon decide the exchange is no longer worth making. For now, these are mostly deliberations, warnings and negotiating positions. The Economist has discussed opting out of Google’s AI features but not exiting Google entirely. Reuters is weighing the economic trade-offs. Politico employees have considered limiting crawler access to free articles. USA Today says it could block Google if the decline continues. Nobody should mistake that caution for weakness. These publishers understand exactly what is at stake. Google made itself indispensable to their audience strategies. Walking away from it alone could turn a deteriorating traffic problem into an immediate traffic collapse. That is precisely why individual action is not enough. ### The Numbers Behind the Anger The Journal used Semrush data to illustrate the decline in Google search traffic, although it later corrected several figures that appeared in its original article and graphic. The corrected June 2025-to-June 2026 comparisons show organic Google search traffic from U.S. users falling 18% for USA Today, 20% for Politico and 31% for both CNN and Business Insider. The corrected trailing 12-month averages show a broader range of outcomes: Publisher | Change in Google search traffic | | The Washington Post | −44% | | Business Insider | −43% | | Reuters | −35% | | USA Today | −28% | | Time | −27% | | The Wall Street Journal | −18% | | People | −3.2% | | BBC | +15% | | The Guardian | +16% | *Source: Corrected Semrush data reported by **The Wall Street Journal**.* The gains at The Guardian and BBC should not be hidden. They also do not disprove the larger argument. The issue is not that every publisher loses the same amount at the same time. The issue is that Google controls the exchange rate between the content it takes and the value it returns—and reserves the right to change that rate whenever it chooses. A publisher that is up 15% today has no contractual guarantee it will not be down 40% next year. Google decides where links appear, when an AI answer appears above them, how prominently sources are cited and whether a user has any reason to leave Google at all. Traffic decline is the symptom. The underlying disease is the concentration of control. ### “Billions of Clicks” is Not an Answer Google says its AI search features send “billions of clicks to the web every week.” That sounds impressive, but aggregate numbers answer the question Google wants asked, not the question publishers need answered. How many clicks would those publishers have received if Google had not placed its own synthesized answer above the links? Google cannot claim credit for every click it still sends without accounting for the clicks its answer engine eliminates. It cannot measure the new bargain only against zero. The comparison is with the referral traffic publishers would have received under the search model Google is actively replacing. Publishers need reporting at the individual-site level. They need to know when their content appears in an AI Overview or AI Mode response, what material was used, how it was attributed, how many people saw it and how many clicked through. In June, Google announced it was beginning to test [new Search Console controls](https://blog.google/products-and-platforms/products/search/new-controls-website-owners/) with a subset of UK publishers. The controls are supposed to allow a site to opt out of helping ground Google’s generative AI search responses without harming its rankings in traditional search. Google also said it would begin supplying more information about which pages appear in AI responses and in which countries. Good. That is movement in the right direction. It is also an admission. Separating traditional search discovery from generative AI consumption was never technically impossible. It was a business choice. Google began testing that separation after pressure from British regulators made its previous position harder to maintain. According to Google, a global rollout is planned only after the UK testing. Publishers should not be grateful that Google is reluctantly testing a measure of control over content Google never created. They should recognize the announcement for what it is: proof that coordinated pressure works. ### The Indispensability Trap This is a near-perfect example of the Indispensability Trap, the subject of my forthcoming book. Google made itself indispensable to publishers by becoming the gateway to discovery. Publishers reorganized their operations around it. They hired SEO teams, adapted headlines, reformatted pages, monitored algorithm changes and accepted Google’s rules because refusing to participate meant becoming invisible. Google then used that dependence to move up the stack. It went from helping users find the answer to becoming the answer. Publishers helped build the platform that now threatens to disintermediate them. They supplied the content that made Google useful, the links that helped establish its authority and the continuous stream of fresh information that kept its index current. Now Google behaves as if no individual publisher is indispensable. Individually, it may be right. Collectively, it is dead wrong. Google can digest yesterday’s web. Its AI can summarize old reporting, combine existing explanations and recycle information that people have already paid to produce. What it cannot do indefinitely is generate tomorrow’s facts. Someone still has to attend the hearing, interview the executive, investigate the company, test the product, analyze the vulnerability, review the legal filing and explain what happened. Someone must finance that work before Google can summarize it. If publishers can no longer support the creation of new information, Google’s answer engine does not become all-knowing. It becomes a sophisticated machine recycling increasingly stale information, derivative reporting and AI-generated content built upon earlier AI-generated content. The parasite eventually destroys the ecosystem that feeds it. ### This Cannot be a Revolt of the Rich The Journal’s reporting naturally focuses on major publishers because those companies have recognizable brands, enormous archives and enough market power to get Google’s attention. But this movement will fail if it belongs only to The New York Times, The Wall Street Journal, Reuters, Reddit, People Inc. and USA Today. The open web is not made valuable solely by its largest publishers. It is made valuable by millions of specialist publications, trade journals, independent journalists, reviewers, researchers, bloggers, local newsrooms and companies that publish expertise directly. Techstrong is part of that ecosystem. Across our properties, we cover cybersecurity, DevOps, cloud-native computing, platform engineering, artificial intelligence, digital transformation and semiconductors. Much of what we publish is too specialized to originate in a general-interest newsroom. Google’s answers appear authoritative partly because they can draw upon publishers that live inside these markets every day. The same is true across healthcare, law, science, finance, education, travel, manufacturing and virtually every other area of human expertise. A small publisher may not have enough leverage to negotiate with Google by itself. Thousands of small and specialized publishers acting under a common standard would represent a body of information Google could not easily replace. This is the unionization moment for the open web. ### A Publisher Compact Publishers should develop a compact built around four principles. First, search indexing must be technically and contractually separated from AI summarization and model training. Allowing Google to help people discover a page is not permission for Google to use that page as a substitute for visiting it. Second, AI consumption must require affirmative consent. Publishers should not be forced to choose between surrendering their content to Google’s AI products and disappearing from traditional search. Third, Google must provide publisher-level transparency. Publishers need to see when their content is used, how it is represented, how often it is cited and what traffic is returned. Fourth, commercial AI use must include attribution and compensation. If Google uses a publisher’s work to make its own product more valuable, that value cannot flow only in one direction. The technical foundation for this kind of action is already being built. Cloudflare launched its [Content Independence Day](https://blog.cloudflare.com/content-independence-day-no-ai-crawl-without-compensation/) initiative around the principle of “no AI crawl without compensation.” A year later, Cloudflare reports that 52% of identified crawler requests are now for AI training, while mixed-use crawlers blending search, agent use and training represent more than 36% of crawler activity. Cloudflare argues that transparency creates control, control creates scarcity and scarcity creates leverage. That is exactly the progression publishers need. Cloudflare also identifies what it calls the “Google convergence problem.” Unlike AI companies that use different crawlers for different purposes, Google combines discovery and AI access in ways that make it difficult for publishers to participate in search without also feeding Google’s AI ecosystem. Google’s new UK control begins to address that problem, but voluntary testing is not enough. Separation should become the worldwide standard. ### Coordination, Not Commercial Suicide Nobody should read this as a demand that every publisher block Google tomorrow morning. That would not be a movement. It would be a collection of individual acts of commercial self-harm. Google’s leverage is real because publisher dependence is real. Smaller sites in particular cannot casually abandon a major source of discovery while hoping direct traffic, newsletters or social media will fill the gap. The goal is coordinated standards, common demands and a credible willingness to withdraw access if Google refuses to negotiate. Publishers need to move together precisely because Google can ignore or punish them when they move alone. The first step is not flipping a switch. It is assembling the coalition, agreeing on the terms and making clear that access to the open web is no longer unconditional. Techstrong should be prepared to join that compact. We are not opposed to search. We are not opposed to AI. We are not asking Google to freeze technology in place or protect every publisher from changes in user behavior. We are defending the economic system that pays people to create the authoritative information on which search and AI depend. Google cannot build its fence around the open web if publishers stop supplying the bricks. It cannot continue moving up the stack if the people producing the raw material collectively insist on a fair exchange. The major publishers quoted by the Journal may be the first to threaten action. They should not be the last. Every publisher, large or small, should be invited to sign on to a common framework of separation, consent, transparency and compensation. How do you kill a parasite? You stop feeding it. One host cannot do that without risking its own survival. But when the hosts act together, the parasite must either change the relationship or starve.