The way the world finds information online is undergoing the most significant structural transformation since Google replaced directory-based web navigation more than two decades ago. The shift has a specific name — the Search Generative Experience — and a specific mechanism: artificial intelligence that synthesizes answers from multiple sources and presents them directly on the search results page, reducing or eliminating the need to click through to any website. For users, this change is often frictionless and convenient. For the publishers, businesses, and creators who depend on search traffic for their livelihood, it is reshaping every assumption about how digital audiences are built. And for the broader information ecosystem, it raises questions about accuracy, attribution, and the economic sustainability of the content creation that AI systems depend on to function.
What the Search Generative Experience Actually Is
Google’s introduction of AI Overviews represents perhaps the most significant change to the search results page in over a decade. Unlike featured snippets, which extract a sentence or two from a single source, AI Overviews synthesize information from multiple sources to create comprehensive answers. This AI-powered feature appears at the top of search results for many informational queries, providing users with detailed responses that often eliminate the need to visit individual websites — ushering in what analysts are calling the zero-click era. Press Information Bureau
The Search Generative Experience was previously tested as an experimental feature in Google Search Labs. On May 14, 2024, Google announced it would be renamed AI Overviews and made available to all US users. It has since expanded globally and is based on Google’s Gemini AI language model.
AI Overviews are built on Google’s generative AI models and deliver an instant synthesis of multiple relevant sources pulled from the web. They consolidate knowledge from multiple sources to present a unified answer — turning Google into both a search engine and an answer engine simultaneously.
The Scale of Adoption
The numbers confirm that this is not a marginal feature. Two billion monthly users engage with AI Overviews globally, up from 1.5 billion in May 2025. AI Overviews were volatile throughout 2025, showing for 6.5% of queries in January, rising to just under 25% in July, and dropping to below 16% in November globally. In early 2025, 88% of queries triggering AI Overviews were informational in nature. By late 2025, AI Overviews started appearing for more commercial queries, increasing from 8% to 18% of commercial searches. Press Information Bureau
AI Overviews currently appear in approximately 40% of informational queries and around 25% of commercial searches. Technology and general knowledge topics see higher rates — in the range of 50 to 60% — while healthcare and finance see lower rates of around 15 to 20% due to YMYL (Your Money or Your Life) content restrictions that require Google to exercise greater caution about AI-generated responses in high-stakes domains. Press Information Bureau
The Zero-Click Consequence: What the Data Shows
The most measurable direct consequence of AI Overviews for the broader internet is the acceleration of zero-click search — queries that are resolved on the search results page without the user visiting any external website. Independent research conducted throughout 2024 and 2025 shows click-through rate reductions ranging from 34% to 46% when AI summaries appear on search results pages. A Pew Research Center study tracking 68,000 real search queries found that users clicked on results 8% of the time when AI summaries appeared, compared to 15% without them — a 46.7% relative reduction in click-through rate.
According to Similarweb data, zero-click searches increased from 56% to 69% between May 2024 and May 2025. Zero-click rates vary further by interface: 34% in Google Search without an AI Overview, 43% with an AI Overview, and 93% in Google’s AI Mode — the experimental full-AI search interface. Press Information Bureau
The effects on specific publishers have been severe in categories where AI Overviews answer questions that previously drove organic search traffic. Learning platform Chegg reported a 49% decline in non-subscriber traffic between January 2024 and January 2025. The decline coincided with AI Overviews answering homework and study questions that previously drove traffic to educational sites. Chegg’s subsequent antitrust lawsuit alleges that Google used content from educational publishers to train AI systems that now compete directly with those publishers. This legal challenge raises a structural question that courts have not yet resolved: whether training AI systems on copyrighted publisher content, and then using those systems to replace traffic to the same publishers, constitutes a compensable harm.
How Search Behavior Is Changing for Users
The user experience of search in 2026 is meaningfully different from 2022 in ways that extend beyond AI Overviews. Conversational search — typing natural questions rather than keyword strings — has become normalized, driven partly by AI Overviews’ ability to handle full-sentence queries effectively, and partly by the growing use of voice search on mobile devices.
AI platforms generated 1.13 billion referral visits in June 2025, representing a 357% increase from June 2024. ChatGPT is the biggest AI traffic referrer, accounting for 50% of AI traffic in recent measurement. AI search traffic converts at 14.2% compared to Google’s 2.8% for traditional search, showing that this audience is dramatically more commercially valuable per visit — likely because users arriving from AI-powered search have already engaged in a more specific, intentional discovery process. Press Information Bureau
AI Overviews are driving over 10% more Google queries for the types of searches that show them, indicating that users actively seek out these AI-powered results rather than being frustrated by them. Press Information Bureau This finding complicates the narrative that AI Overviews are universally harmful — they appear to be increasing the overall volume of searches in the categories where they appear, even as they reduce per-search click-through rates.
The New Rules of Content Visibility: GEO Replaces SEO
For businesses, creators, and publishers, the response to this environment has produced a new discipline: Generative Engine Optimization, or GEO. Where traditional SEO focused on ranking in position one through ten on the search results page, GEO focuses on being cited as a source within AI-generated summaries. The two objectives overlap but are not identical — a website can rank on page one without being cited in an AI Overview, and can be cited in an AI Overview without ranking prominently in traditional results.
Success in the new search landscape is no longer just about ranking on page one. It is about being credible and “source-worthy” within AI-generated summaries. Brands must push toward AI-first SEO strategies, address significant drops in organic click-through rates, and adapt to generative search optimization.
Google continues to prioritize content that aligns with E-E-A-T principles — Experience, Expertise, Authoritativeness, Trustworthiness. With SGE, this becomes even more important: only credible, well-researched, and value-driven content will be considered as a source for AI-generated overviews. As SGE utilizes natural language processing, search queries have become longer and more conversational, making optimization for long-tail, question-format queries more valuable than targeting short competitive keywords. Structured data markup — FAQ, How-To, Article, Review schemas — helps Google’s AI categorize and surface content in generated responses.
As the search results page becomes more generative, SEO is becoming more about presence than position. The question is no longer only “Can I rank for this keyword?” but “Will an AI system cite me as a trustworthy source when answering this question?”
The Accuracy Problem: When AI Answers Are Wrong
The rollout of AI Overviews has not been without challenges. Google faced criticism for inaccurate information and unusual suggestions in early iterations, leading to refinements in how the system evaluates and presents information. Press Information Bureau The specific failures in the early rollout of SGE — AI Overviews that suggested adding glue to pizza sauce, or recommended eating one small rock per day for health benefits, examples widely documented in media coverage — were corrected quickly by Google but exposed a fundamental vulnerability in systems that synthesize information without contextual judgment.
Google has since implemented more cautious handling of YMYL queries — medical, legal, and financial questions where inaccurate AI-generated answers carry real-world consequences. The long-term question of how reliably an AI system can distinguish authoritative from unreliable sourcing, across the full breadth of human inquiry at two billion queries per day, is one that researchers are studying and that no current AI system has fully solved.
For ordinary users, the practical implication is clear: AI-generated search summaries are useful tools for initial orientation, but any information that will inform a consequential decision — medical, financial, legal, or otherwise — should be verified against primary sources before being acted upon.