Publishers vs. Google: What the Conflict Over AI Search Means for Armenian Media
YEREVAN, July 30. /ARKA/. Just a few years ago, even posing the question seemed almost impossible: why would a media outlet voluntarily close itself off from the world's largest search engine?
Now this option is being seriously discussed by USA Today, Reuters, Politico, People Inc., and other major media companies. The online platform Reddit is also reconsidering its relationship with Google.
The reason is simple: the old model of interaction between media and the search platform is beginning to break down.
Previously, it looked something like this: a publisher allows Google to index its content, and Google drives readers to the publisher's website. The media receives an audience, advertising revenue, and subscribers.
With the advent of AI search, this model is changing. Google is increasingly generating ready-made responses based on materials from the open internet, and users no longer need to go to the original source.
Again, search traffic for a number of major media outlets is noticeably declining. However, the data itself requires careful interpretation: traffic is influenced not only by AI features, but also by changes in search algorithms, seasonality, editorial strategy, and audience behavior.
According to revised Semrush data cited by The Wall Street Journal, from June 2025 to June 2026, organic search traffic from Google from US users decreased by 18% for USA Today, by 20% for Politico, and by 31% for CNN and Business Insider.
The picture is mixed, however. According to the WSJ chart, average traffic over the past 12 months increased by approximately 16% for The Guardian and by 15% for the BBC.
The scale of the change is also demonstrated by Cloudflare statistics published on July 1, 2026, in its analytical report "Content Independence Day, one year on." According to Cloudflare, more than half of the internet traffic passing through its network is now generated by automated systems. Moreover, 52% of crawler requests in June 2026 were for data collection for AI training, compared to 22% in spring 2025.
However, this refers specifically to the number of requests, not the proportion of bots among website visitors. A single bot can send thousands of requests in a short period of time, so these statistics cannot be directly compared to the number of real readers.
But the most interesting thing isn't even the numbers.
Publishers are starting to ask a different question: why should participation in organic search simultaneously mean consenting to the use of content in AI Overviews and other Google generative features?
It's important to distinguish between two mechanisms. Google-Extended allows publishers to restrict the use of content for training Gemini and certain other AI systems without abandoning organic search. However, until recently, it was virtually impossible to separately prohibit the use of materials in AI Overviews and AI Mode while maintaining traditional search indexing.
On June 3, Google announced the start of testing in the UK of a new tool in Search Console that allows users to opt out of generative search features without losing visibility in organic search results. The testing began as part of the UK regulator's compliance with the Competition and Markets Authority (CMA), which previously granted Google strategic market status in general search and search advertising. The company intends to roll out this mechanism globally after the testing, but no specific timeframe has yet been announced. Meanwhile, publishers are considering various scenarios.
USA Today is considering a complete Google block if the decline in search traffic continues. USA Today Co. CEO Mike Reed stated: "It's time to take a stand and say: enough!"
Notably, the company is already suing Google over its alleged monopoly in advertising technology, meaning the publisher has experience in legal battles with the corporation.
Reuters is exploring the possibility of limiting the scanning of its open consumer news products.
Politico is discussing restrictions on bots and introducing mandatory user registration.
The Economist is considering abandoning AI search while maintaining a presence in traditional search results.
People Inc. is also not ruling out a complete Google block.
Reddit is renegotiating a separate licensing agreement with Google, worth approximately $60 million per year, which allows the company to use the platform's materials to train AI models.
So, there's no talk yet of a mass exodus of media outlets from Google. A more important process is underway: the largest publishers are reconsidering the very economics of their relationship with the search platform.
Google denies that its AI features are simply stealing traffic from websites. The company claims that its AI search features drive billions of clicks to websites weekly, help publishers find new audiences, and provide links to primary sources.
However, Google does not disclose aggregated data that would allow for an independent assessment of the distribution of these clicks among publishers and compare them with previous search traffic.
Therefore, the question, in my view, is now very simple: if a media outlet hands over its content, data, and expertise to a platform, what economic value does the platform return in return?
Previously, the answer was traffic. In the age of AI, that answer is becoming increasingly less clear. I believe that the distribution of this value will be one of the main conflicts between major media and Big Tech in the coming years.
What does this mean for Armenian media?
The situation is more complex for Armenian media than for major Western publications. They have fewer legal and negotiating leverage against Google. Therefore, reducing dependence on any single external platform becomes their primary tool.
Avoid single-channel dependence. Some large publishers entered the AI era with a high dependence on search traffic. Armenian media need to assess their dependence on Google, Facebook, Telegram, YouTube, and other channels based on their analytics and avoid critical audience concentration on a single platform.
Diversify not only channels but also formats. Videos, podcasts, live broadcasts, interactive projects, and events don't offer complete protection from AI search, but they create additional ways to attract and monetize audiences beyond simple Google text clicks.
Build direct relationships with readers. Email newsletters, messenger channels, mobile apps, subscriptions, and loyalty programs allow media outlets to be less dependent on external algorithms. For a small market, this is more realistic than independently applying legal or negotiating pressure to a global corporation.
Monitor changes rather than making hasty decisions. At this stage, there is no reason to recommend that Armenian media completely shut themselves off from Google. Organic search traffic, AI-driven traffic, and the share of direct visits should be separately monitored.
Invest in unique local content. General information materials, which can be compiled from a variety of open sources, are likely the most vulnerable to AI search substitution. Exclusive news, proprietary data, local expertise, interviews, investigations, and deep knowledge of Armenian specifics are more difficult to replace with automated responses.
The main conclusion for Armenian media is not to immediately block Google. It is to begin building a model now in which neither Google nor any other technology platform can single-handedly determine the size of the audience and the economic sustainability of the editorial staff.
Konstantin Petrosov
Director of ARKA News Agency
