Google’s AI Search Is Squeezing Publishers — and Forcing a Reckoning Over Traffic

Google’s AI search features have cut into publisher traffic, weakened the value of traditional search referrals, and forced news organisations to rethink how they reach audiences since AI Overviews launched in 2024 and AI Mode arrived in May 2025. The shift matters because it moves attention away from publisher sites and toward Google’s own answers, changing the economics of digital journalism.

What changed in Google Search?

Google did not simply add a new box to search results. It changed the experience of search itself. With AI Overviews, launched in 2024, Google began placing machine-generated summaries above standard links. With AI Mode, rolled out in May 2025, Google went further and replaced the familiar results page with a chatbot-style interface.

For publishers, those changes altered the basic bargain of search. In the old model, a query could lead to a headline, a click, and a visit that a news site could monetise through advertising, subscriptions, or membership upsells. In the new model, a user can get an answer without ever leaving Google.

The Verge’s reporting shows that publishers quickly noticed the consequences. Media outlets said AI Overviews threatened both traffic and visibility, while industry groups argued the product was stripping away the incentive to visit source sites at all.

Why it matters

The stakes are bigger than one product update. Search has long been one of the most important entry points to journalism on the open web, especially for publishers that do not have huge direct-audience habits like home-page visits or loyal app users.

When search clicks fall, the pressure spreads across the business:

• ad impressions can fall with traffic;
• subscription funnels can narrow;
• newsletter sign-ups can slow;
• brand visibility can weaken if a publisher no longer appears as a destination.

That is why publishers have framed Google’s AI search shift as more than a product change. The News/Media Alliance warned the change would be

catastrophic

for web traffic, according to The Verge, reflecting the concern that Google is keeping more attention inside its own interface while sending less of it to the publishers that supply the underlying reporting.

The broader issue is who captures value from information. Publishers create the articles, interviews, and reporting that feed search results. Google now increasingly packages that information into its own answer layer. If users stop clicking, the publisher may do the work while Google keeps the audience relationship.

How quickly did the impact show up?

The early signs appeared fast. Press Gazette found AI Overviews on almost a quarter of news-related searches in May 2024, a striking share so soon after launch. Later, Press Gazette reported that AI Overviews were replacing media sites that had previously held top search positions.

Those are not abstract changes buried in analytics dashboards. They show a visible reshuffling of search real estate, with AI summaries occupying the most prominent space and traditional links pushed lower.

That pattern has been echoed in data points from major media companies and industry trackers. The Wall Street Journal reported traffic declines at Business Insider, The Washington Post, HuffPost, and The Wall Street Journal itself. Similarweb data showed New York Times search traffic falling from 44% to 35.5% over three years. Digiday reported that 19 Digital Content Next members lost Google traffic between May and June 2025, with declines ranging from 1% to 25%.

In the United Kingdom, DMG Media told the UK Competition and Markets Authority that click-through rates dropped by up to 89% between June and July 2025. That figure, if sustained, points to a severe collapse in the traditional search referral model.

By the numbers

What publishers say is happening to their traffic

Publishers’ complaints are consistent across the fact pattern in The Verge’s reporting: the search result is no longer a doorway, but a destination in itself.

That shift helps explain why some outlets have seen not just traffic pressure but a change in the quality of traffic they receive. Google has said AI search sends more clicks to sites with forums, videos, podcasts, and user posts, suggesting the company believes some forms of content can still win audience in a richer search environment. But for traditional news publishers, that is cold comfort if the top answer is now generated by Google before a user ever reaches the article.

Google Search head Liz Reid has said click volume stayed

relatively stable

, according to The Verge. Google senior vice president of knowledge and information Nick Fox has also defended the product, saying

the web is thriving

. Those statements cut directly against the publishers’ argument that Google is starving the open web of referral traffic.

The dispute is not only about whether clicks exist, but where they go, how valuable they are, and who benefits when searchers get enough information from the summary alone.

Background: how we got here

For years, Google Search worked as a giant distributor of web traffic. A search query produced links, and those links delivered readers to publishers’ sites. That arrangement helped build the modern digital news economy, even as it also made publishers dependent on a single platform they did not control.

AI Overviews changed that formula by putting a synthesized answer above the links. AI Mode went further by making search feel like a chat session rather than a list of destinations. Together, the products reduce the need for a user to leave Google at all.

The tension surfaced quickly because publishers were already under pressure from other platform shifts, changing audience habits, and ad-market instability. Search was one of the last large, measurable sources of free audience flow. If that source weakens, publishers must compensate elsewhere.

That is why the industry response has been so broad. Some publishers have sued Google over AI search traffic losses. Judge Amit Mehta dismissed the antitrust case brought by Chegg and Penske Media, and separately rejected an Arkansas newspapers class action over AI Overviews. Those rulings do not end the debate, but they show how difficult it can be for publishers to turn grievance into legal relief.

What Google says in response

Google’s message has been that search remains healthy and that its AI tools still send users outward. The company has said click volume is broadly steady and that AI search helps direct traffic to certain kinds of sites, including forums, video, podcast and user-post platforms.

The company has also begun making changes that appear designed to address some of the criticism. After a UK order, Google rolled out an opt-out toggle for AI Mode and AI Overviews. It is also showing sources more prominently in both products. In addition, Google is testing payments to publishers for content used in AI Overviews and AI Mode.

Those steps matter, but they do not settle the larger question. Better attribution and source labels may help users identify where information comes from. They do not necessarily restore the click volume publishers say they have lost.

How are publishers adapting?

Publishers are not waiting for search to return to its old shape. The reporting indicates they are leaning more heavily on social media, newsletters, and subscriptions to compensate for weaker Google referrals.

Three major adjustments are emerging

1. Direct audience building: newsletters, apps, and membership products aim to reduce dependence on search.
2. Platform diversification: publishers are pushing harder on social channels and other discovery routes.
3. Revenue rebalancing: subscription and recurring-reader income becomes more important when ad-supported traffic softens.

This is not a clean pivot. Most publishers still need search, and not all traffic can be replaced with loyal direct users. But the direction is clear: the more Google answers the question itself, the more publishers have to chase readers elsewhere.

What happens next?

The next phase will depend on whether Google’s latest changes actually preserve referral traffic, or merely soften the appearance of the shift.

Watch four questions in particular:

• Whether source prominence helps. Google is showing sources more prominently, but it is not yet clear whether that changes behaviour at scale.
• Whether payments expand. Google is testing payments to publishers in a pilot that includes 400 publishers, but the commercial terms are not public here.
• Whether more publishers resist. Some outlets may try to block indexing or limit AI use of their content if traffic erosion continues.
• How large the revenue hit becomes. The long-term impact on publishing businesses will depend on whether traffic declines are temporary, concentrated, or structural.

The larger strategic issue is whether Google’s AI search becomes a more closed information system, with the platform absorbing more of the reader relationship and the publishers receiving less in return. If that happens, the search industry may remain healthy while the open web around it becomes thinner.

Timeline of the shift

The bigger question for the open web

The story here is not only about Google and publishers. It is about the future economics of finding information online.

If users increasingly trust the answer that appears on the results page, then the value of the original source changes. For news publishers, that can mean fewer readers, less leverage, and a greater need to build audience loyalty off-platform. For Google, it means a more powerful and more self-contained search product. And for readers, it means a more convenient answer — but often one that arrives with less visible connection to the reporting behind it.

That is the central tension shaping this moment: Google’s AI search may be improving the search experience on its own terms while shifting the costs onto publishers that depend on the open web for their business models.