AI Overviews, Google's AI-generated search summaries, became a core feature of Search within a year of launch. Their share of search appearances expanded from 15% last year to 43% this year.
Close to a threefold increase in twelve months — meaning the odds of encountering an AI summary on a results page now approach one in two.
Visits to the related AI Mode surged as well:
More than double in a year.
The behavioural change matters more than the volume. Users are entering long natural-language questions rather than short keywords, and expanding information conversationally within Google.
This undermines a core assumption of keyword-led SEO. When the input shifts from "running shoe recommendations" to "what running shoes work for someone with flat feet and bad knees," the kind of content that matches changes too. Long-tail questions that never surface in keyword volume data multiply.
Publishers are worried about traffic decline from cases where AI summarises website content without generating a click.
Citation data explains the concern. Source attribution in AI responses increased more than fivefold over the past year — yet only 6.8% of ChatGPT desktop responses included a source link.
The growth rate looks like progress; the absolute figure remains negligible. A fivefold increase producing 6.8% summarises the current state.
Hence the analysis that Google is becoming search's destination rather than a waypoint. What was once a stop on the way to another site is increasingly the endpoint itself.
Stop judging content performance on traffic alone. A pattern of stable impressions with falling clicks likely means absorption into AI summaries. Tracking impressions and clicks separately in Search Console is what makes the cause visible.
Separate being cited from being named. A source link without your brand name in the answer text leaves the reader unaware of you. That gap is covered in Ghost Citations: AI Search Uses Your Content Without Naming Your Brand 40% of the Time.
Restructure content around natural-language questions. Using actual question sentences as subheadings rather than keyword headings positions content better for conversational search — one reason FAQ structure matters again.
Fix your measurement set. Without deciding what AI search visibility means for you, improvement is unmeasurable. "Do We Show Up in ChatGPT?" — Four Metrics That Answer 2026's Top Client Request is a useful framework.
For how the players in Korea's search market are dividing along different axes, see Naver vs Google vs Daum: Korea's AI Search Rivalry Splits Into Two Camps.
Their share of search appearances rose from 15% last year to 43% this year — close to a threefold increase within a year of launch.
From 126 million in June 2025 to 279 million in May 2026, more than doubling in a year.
Users are entering long natural-language questions instead of short keywords and expanding information conversationally within Google.
Source attribution increased more than fivefold over the past year, but only 6.8% of ChatGPT desktop responses included a source link — supporting the view that Google is becoming search’s destination.
To apply what you just read to your own site, start with a free audit of where things are now.
A strategist replies within 24 hours on business days.
An E-E-A-T checker was never automatable because there is no API for it. Here is how an AI coding agent read Google's quality guidance, crawled a page sample, and produced a scored audit — plus where it stops.
A GEO audit is not checking whether your brand appears in an AI answer. Here is a nine-point GEO checklist split into four on-site items, four off-site, and one measurement layer.
LinkedIn, YouTube, Reddit, and TikTok are building systems to identify low-value AI content. Here are the limits of watermarking and what content still gets distributed.