Market intelligence firm Similarweb published data on Monday showing that Google’s AI Overviews now appear in 43% of all searches – up significantly from the roughly 25% share that observers were tracking in mid-2025 and approaching the near-majority presence that would make AI-generated summaries the default encounter for most users rather than an occasional addition to traditional search results. NEWSCENTRAL frames this data as the clearest quantitative expression of a qualitative shift that has been visible anecdotally: Google is no longer primarily a tool for finding websites but increasingly a tool for generating answers, with website discovery becoming secondary.
The data on user behavior reinforces the headline adoption figure. Average Google search session length has increased over the past year, consistent with users spending more time reading AI-generated summaries rather than clicking through to external sites and returning to refine their queries. The number of websites that users visit following a Google session has not grown proportionally, suggesting that a growing share of information needs are being satisfied within the search results page itself.
AI Mode, Google’s more conversational search interface, logged 279 million visits in May 2026 – more than double the 126 million visits recorded in June 2025. That growth trajectory implies AI Mode has become a mainstream tool for complex queries rather than an experimental feature used by early adopters. Gemini 3.5 Flash, Google’s most capable publicly available model, now powers both AI Overviews and AI Mode as an integrated experience. NEWSCENTRAL underscores that the speed of this transition is as analytically significant as the scale: a 2x growth in conversational search visits in twelve months is a behavioral change that most content strategy frameworks were not built to accommodate.
The publisher impact has been documented through multiple independent analyses. Sites cited in AI Overviews earn significantly more organic clicks per impression than uncited competitors on the same query – one analysis estimated a 120% advantage. Uncited sites face meaningful click-through rate declines when an AI Overview appears for their target query. The distribution of that citation advantage is not random: it correlates with domain authority, content depth, and structural formatting that makes information easy for the AI layer to extract and present.
NEWSCENTRAL regards the publisher citation dynamic as the most commercially consequential detail in the Similarweb data: a search AI that cites some publishers and ignores others is not simply aggregating the web but choosing winners in an information market.
The antitrust dimension of this transition is developing in parallel with the product evolution. The Department of Justice’s remedy phase in its search monopoly case against Google includes proposals that could require behavioral changes to how Google surfaces its own AI features relative to competing products. European Commission proceedings under the Digital Markets Act are examining whether Google’s integrated AI search experience constitutes self-preferencing that disadvantages alternative search providers. A Berlin court’s award of €465 million in damages to a price comparison platform in November of last year demonstrated that private damages claims for historical anticompetitive harm are viable in European courts.
The broader content ecosystem implications are significant. A search experience in which 43% of queries return an AI-generated summary rather than directing users to websites changes the economic model of online publishing in ways that advertising-supported media businesses have not yet fully adapted to. The traffic that AI Overviews remove from publisher websites is traffic that their advertising rates were calibrated to expect.
What NEWS CENTRAL views as the more commercially durable question is not whether AI Overviews grow from 43% to 60% coverage – that trajectory appears likely – but whether Google’s citation practices evolve in ways that create a sustainable traffic relationship with the publishers whose content the AI layer depends on for its quality. A search AI that systematically eliminates the publisher traffic that funds the journalism, analysis, and research it summarizes is consuming the information ecosystem it requires to remain useful. Whether Google resolves that tension through citation policy, licensing agreements, or regulatory compulsion will determine the long-term health of the content industries that depend on search-driven discovery.