One of the oldest and most important understandings of the internet was very simple. Publishers would let Google crawl and index their content, and Google would send readers of that content to their websites. News outlets, databases, small businesses and various online services would sustain that traffic through advertising, subscriptions and direct reader relationships. But with the spread of AI-driven search, that old system now faces major change. An opinion piece by Salih Kaya, published in Daily Sabah, raises exactly this concern.
Google’s AI Overviews now present a summarized answer to a user’s question before sending them to any website. As a result, users often don’t need to visit the relevant website to learn something. Google’s new AI Mode goes a step further, enabling conversational search, information synthesis and completing various tasks all on the same platform. In other words, Google is trying to become not just a guide to the web, but the primary destination for finding information itself.
Some statistics on its impact have already raised concern. Chartbeat data shows that by the end of 2025, traffic to various news websites from Google search had fallen by nearly a third compared with the previous year. For U.S. publishers, that decline was around 38%. The same article cites Sparktoro figures showing that in 2026, about 68% of Google searches end without the user clicking through to any outside website. In 2019, that figure was 49%.
This kind of search is called “zero-click search.” That is, the user asks a question but gets the answer right on the search page itself, with no need to visit a separate website. A recent UCLA analysis also shows that more than two-thirds of Google searches in the U.S. now end without any click to a public website. This means that as the flow of information on the internet increases, readers’ direct access to the original source websites of that information may actually be declining.
The impact of declining clicks falls hardest on organizations that depend on search-driven traffic. Citing research from Seer Interactive, the Daily Sabah article notes that since AI Overviews rolled out broadly, both organic and paid click-through rates have dropped sharply. Even when an AI answer includes a link to a website, the rate of clicks on that link has fallen significantly.
As a result, small and mid-sized news outlets may be at the greatest risk. Major news brands have direct readers, subscribers and followers built through social media. But local-language media, specialized websites and new publications largely depend on search engines to reach new readers. If their search visibility declines, their advertising revenue, subscriptions and ability to build readership could shrink as well. This raises the risk of damage to media pluralism overall.
This is where the core conflict lies. Google’s AI models rely on the vast amount of information published across the web to generate answers. But when that answer is delivered directly on Google’s own platform, the publishers who created the original content lose readers. In other words, the websites that produce information allow AI search to generate more effective answers using their content — yet that very answer can free the user from ever needing to visit the original website. Many publishers see this as a “structural crisis in traffic.”
Google, for its part, hasn’t fully accepted this criticism. The company claims AI Overviews and AI Mode are actually trying to make website links more visible. In 2026, Google added inline links, website previews and a “preferred sources” feature. Users can mark their preferred news outlets as preferred sources. According to Google’s own figures, AI Overviews now has more than 2.5 billion monthly active users, and AI Mode has more than 1 billion monthly users.
But this very move by Google raises another question: if the user experience in AI search changes so much that most information can be found within Google itself, how will the original web publishers’ economic model survive? There is a big difference between simply showing a link and actually sending a reader. For a website, real readers, page views, ad impressions and subscriptions matter far more than thousands of mere impressions.
For this reason, the issue is no longer just a dispute between tech companies and news media; it has also become a question of competition policy and the digital economy. The European Commission is investigating the use of publishers’ content in Google’s AI search. European regulators are asking publishers whether the option to opt out of allowing their content to be used in Google’s new AI search system is sufficient for them.
Since Aug. 31, Google has rolled out a control for publishers worldwide that lets them stop their content from being used in AI Overviews and AI Mode if they choose. However, an important condition applies: opting out of AI search also means the site will get no traffic or impressions from the AI features. Google says this decision will not affect a site’s ranking in regular Google search.
This change is part of a larger transformation in Google’s business model. Google is gradually moving toward an “agent manager” system, in which the search engine won’t just send users to sources of information, but will itself gather information, analyze it, generate answers and complete various tasks. This will keep users inside Google for longer, reducing the need to visit outside websites.
So this change should not simply be seen as “Google versus publishers.” It is tied to the future economics of the internet itself. If websites keep producing information in the days ahead, but the economic benefit of that information goes mainly to AI platforms, the incentive to create new, independent content could shrink. This could become a major challenge, particularly for investigative journalism, local news, specialized research and long-term, information-based content.
Still, a ban is not the only path against AI search. Publishers, tech companies and regulators need to build new frameworks for fairness, transparency, copyright, licensing and traffic-sharing around the use of content. At the same time, news media also need to build direct readers, subscriptions, memberships, newsletters and their own digital communities, rather than relying solely on search.
Because Google’s AI search brings a change that is perhaps not primarily technological — it is reshaping the internet’s power structure. The search engine was once the door to the web; now, if the full answer can be found inside that door itself, the question arises — what is the future of the websites that lie beyond it?
That’s why viewing Google’s AI Overviews merely as a convenience for quickly finding information misses the full picture. It could simultaneously shape the flow of information, the economics of journalism, the incentive to create content, and the future of the open web. If AI makes the web richer, that will be a triumph of technology. But if it uses the web’s content to strip away websites’ readers and revenue, then the very open internet system that helped AI grow could one day become its greatest source of harm.