AI Overviews Cut Site Clicks as Google Expands AI Everywhere

AI Overviews Put Publishers on the Defensive

Site owners and SEO practitioners are increasingly asking a pointed question: is Google's AI Overviews feature siphoning traffic away from their websites, and if so, what can be done about it? Industry guidance now walks publishers through using Search Console data to detect click declines tied to AI-generated summaries and to identify which keywords are still worth fighting to rank for 1. The advice reflects a broader anxiety in the publishing world that as Google answers more queries directly on the results page, fewer users ever click through to the original source.

That concern is compounded by Google's parallel push toward "Preferred Sources," a mechanism reportedly reshaping how content is selected and surfaced within AI Overviews. Commentary on this shift frames it as forcing publishers to rethink content strategy entirely, since simply ranking well in traditional blue-link results no longer guarantees visibility or traffic when an AI-generated summary can answer a query outright 5. Together, these developments suggest SEO is moving from a discipline focused on rankings to one focused on survival within AI-mediated search experiences.

A Broader AI Push Across Google's Ecosystem

The pressure on publishers is unfolding against the backdrop of Google's aggressive expansion of AI across its entire product line. At a recent event, Google unveiled new Pixel phones featuring slimmer camera bumps, more powerful zoom capabilities, and deeper AI integration designed to let users complete tasks with fewer taps and swipes 26. Coverage of the launch emphasized that Google has been steadily layering more AI into its phones since 2023, with the new devices representing the latest step in that trajectory 78. Pricing details accompanied the announcement, with reporting noting Google's clear commercial motive: giving consumers a reason to upgrade while growing the base of people actively using its AI tools 4.

That strategy extends to the underlying models themselves. Google and OpenAI have both introduced fast, low-cost AI models aimed at powering cheap, responsive agents — Google's Gemini 3.7 Flash is already publicly available, while OpenAI's competing GPT-5.6 Sol Ultrafast remains restricted to an invite-only waitlist 3. This model race matters because faster, cheaper inference is precisely what makes features like AI Overviews economically viable at Google's scale, reinforcing the likelihood that AI-generated answers will keep expanding across search results.

Why It Matters

Taken together, the coverage paints a picture of an AI ecosystem tightening around Google's control: phones, models, and search results are all being optimized to keep users inside Google's AI layer rather than routing them to outside websites. For publishers, that means the fight over Search Console clicks isn't an isolated SEO problem — it's a symptom of a much larger shift in how Google monetizes and mediates information across every product it ships.