Google AI Overview
An AI-generated summary feature introduced by Google Search, presented at the top of the search results page as a synthesized answer produced by AI. It cites web page sources within the summary.
Core Facts
Timeline (last 90 days)
AI Overview triggering and citation are two separate stages: triggering is determining whether to generate an AI summary, and citation is selecting source pages after generation.
BrightEdge's analysis set the overlap rate between AI Overview citations and organic search rankings at 17%.
WashU research measurements show that 13.7% of all searches triggered an AI Overview.
Queries in question form triggered an AI Overview 64.7% of the time, while non-question queries triggered it only 9.5% of the time.
Topic type affects AI Overview trigger rates: hobbies and leisure 46.1%, science 39.9%, politics only 7.5%, law and government 9.6%.
After scoring the domain credibility of 61,212 citations, the study found that the domains cited by AI Overviews had a higher average credibility score than the traditional first-page results shown in the same search.
Each AI Overview cites about eight sources on average, with the top 5 hostnames accounting for 20% of citations, the top 50 accounting for 48.3%, and the remaining 43% spread across 7,379 hostnames.
The study broke down AI Overview answers into 98,020 atomic claims, of which 89% were explicitly or roughly supported, 7% were not mentioned in the cited text, and 4% directly contradicted the cited sources.
Only 41.9% of AI Overviews were fully grounded, meaning every claim had support.
The query fan-out mechanism means that whether a page can be cited by an AI Overview depends on whether it matches one of the fan-out subqueries, not on its ranking for the original main query.
12 more timeline events
All Facts (20)
According to data from research institutions, the emergence of AI Overviews has caused the click-through rate of traditional organic search results to decline by approximately 30% on average.
65%UnverifiedThe underlying architecture of AI search products such as Perplexity and Google AI Overview is based on the RAG (Retrieval-Augmented Generation) mechanism.
60%UnverifiedAI Overview triggering and citation are two separate stages: triggering is determining whether to generate an AI summary, and citation is selecting source pages after generation.
50%UnverifiedQueries in question form triggered an AI Overview 64.7% of the time, while non-question queries triggered it only 9.5% of the time.
50%UnverifiedTopic type affects AI Overview trigger rates: hobbies and leisure 46.1%, science 39.9%, politics only 7.5%, law and government 9.6%.
50%UnverifiedAfter scoring the domain credibility of 61,212 citations, the study found that the domains cited by AI Overviews had a higher average credibility score than the traditional first-page results shown in the same search.
50%UnverifiedEach AI Overview cites about eight sources on average, with the top 5 hostnames accounting for 20% of citations, the top 50 accounting for 48.3%, and the remaining 43% spread across 7,379 hostnames.
50%UnverifiedThe study broke down AI Overview answers into 98,020 atomic claims, of which 89% were explicitly or roughly supported, 7% were not mentioned in the cited text, and 4% directly contradicted the cited sources.
50%UnverifiedOnly 41.9% of AI Overviews were fully grounded, meaning every claim had support.
50%UnverifiedThe query fan-out mechanism means that whether a page can be cited by an AI Overview depends on whether it matches one of the fan-out subqueries, not on its ranking for the original main query.
50%UnverifiedA study analyzing 55,393 Google searches found that 29.8% of domains cited by AI Overviews did not appear in the first page of search results at all.
50%UnverifiedAI Overview uses a query fan-out mechanism, splitting the original query into multiple subqueries that are retrieved separately and then assembled into an answer.
50%UnverifiedAccording to Ahrefs' analysis of 863,000 keywords and 4 million AI Overview URLs, only 38% of cited pages appeared in the top 10 organic search results, whereas this proportion was still 76% in July 2025.
50%UnverifiedBrightEdge's analysis set the overlap rate between AI Overview citations and organic search rankings at 17%.
50%UnverifiedWashU research measurements show that 13.7% of all searches triggered an AI Overview.
50%UnverifiedWhen users ask ChatGPT, Perplexity or Google AI Overview questions, the way brands are mentioned, cited or recommended is completely different from traditional search engines.
50%UnverifiedTraditional SEO is being reshaped by AI search; when users ask ChatGPT, Perplexity or Google AI Overview questions, what they get is an answer directly generated by the LLM rather than a list of links.
50%UnverifiedThe user considers the response quality after switching to be very poor, almost worse than Google AI Overview.
50%UnverifiedAI Overview leverages RAG (Retrieval-Augmented Generation) technology combined with the Gemini series of large language models for content synthesis.
50%UnverifiedIn actual development, fallback logic should be added to AI visibility tools to handle situations where no AI Overview results are returned.
50%