{"report":"The State of AI Search 2026","version":"2026.1","evidenceCutoff":"2026-08-13","nextScheduledReview":"2027-01-15","method":"VITON13 separates observed behavior, self-reported behavior, independent audits, platform capabilities, operator claims, vendor analytics and editorial inference. Values from unlike populations are not combined into a market-share estimate.","evidenceClasses":{"observed_behavior":"Behavior recorded from opted-in browsing or product analytics.","self_report":"What respondents said in a defined survey population.","independent_audit":"A third party tested system outputs against a declared rubric.","product_capability":"A feature documented by the product operator.","operator_claim":"A scale or performance statement made by the product operator and not independently audited here.","vendor_analytics":"Analytics published by a commercial measurement vendor using its own customer or panel data.","viton13_analysis":"A VITON13 interpretation derived from cited evidence, not a measured population statistic."},"records":[{"id":"DNR25-01","evidenceClass":"self_report","claim":"Weekly use of generative AI across six surveyed markets rose from 18% in 2024 to 34% in 2025.","population":"Adults in Argentina, Denmark, France, Japan, the United Kingdom and the United States","period":"2024-2025","source":"Reuters Institute, Generative AI and News Report 2025","url":"https://reutersinstitute.politics.ox.ac.uk/generative-ai-and-news-report-2025-how-people-think-about-ais-role-journalism-and-society","caveat":"Six-market self-report; it is not a global usage or search-market-share measure."},{"id":"DNR26-01","evidenceClass":"self_report","claim":"Weekly use of standalone AI chatbots for news increased from 7% to 10% across the 2026 Digital News Report markets.","population":"Almost 100,000 online news consumers across 48 markets","period":"2025-2026","source":"Reuters Institute, Digital News Report 2026","url":"https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2026/dnr-executive-summary","caveat":"News use only; adoption varied materially by country."},{"id":"DNR26-02","evidenceClass":"self_report","claim":"Sixteen percent of under-35s reported weekly AI-chatbot use for news, while only 1% of all respondents named AI as their main news source.","population":"Digital News Report 2026 respondents across 48 markets","period":"2026","source":"Reuters Institute, emerging uses of AI chatbots for news","url":"https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2026/emerging-uses-ai-chatbots-news-and-what-it-means-journalism","caveat":"The two percentages have different bases and describe news, not every information task."},{"id":"DNR26-03","evidenceClass":"self_report","claim":"Forty-two percent of AI-chatbot news users said the ability to ask follow-up questions was a valued feature.","population":"AI-chatbot news users in the Digital News Report 2026","period":"2026","source":"Reuters Institute, Digital News Report 2026 executive summary","url":"https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2026/dnr-executive-summary","caveat":"This measures stated preference among users, not the frequency or quality of follow-up sessions."},{"id":"PEW25-01","evidenceClass":"observed_behavior","claim":"Eighteen percent of tracked Google searches produced an AI summary, and 58% of panelists encountered at least one during March 2025.","population":"900 opted-in United States adults with tracked browsing","period":"March 2025","source":"Pew Research Center","url":"https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/","caveat":"One country, one search engine and one month; exposure is not adoption."},{"id":"PEW25-02","evidenceClass":"observed_behavior","claim":"Panelists clicked a traditional result after 8% of Google visits with an AI summary versus 15% without one; only 1% clicked a cited source in the summary.","population":"900 opted-in United States adults with tracked browsing","period":"March 2025","source":"Pew Research Center","url":"https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/","caveat":"Observational association, not a randomized causal estimate; click definitions were limited to the tracked next visit."},{"id":"PEW25-03","evidenceClass":"observed_behavior","claim":"AI summaries appeared for 53% of tracked queries with ten or more words, compared with 8% of one- or two-word queries.","population":"Tracked Google searches from the same United States panel","period":"March 2025","source":"Pew Research Center","url":"https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/","caveat":"Query length correlates with other properties such as question format and complexity."},{"id":"TOW25-01","evidenceClass":"independent_audit","claim":"The Tow Center ran 1,600 excerpt-identification queries across eight generative search products and documented incorrect, incomplete and fabricated citations.","population":"20 publishers, 10 articles per publisher and eight products","period":"2025","source":"Tow Center for Digital Journalism / Columbia Journalism Review","url":"https://www.cjr.org/tow_center/we-compared-eight-ai-search-engines-theyre-all-bad-at-citing-news.php","caveat":"A targeted news-retrieval test; it does not estimate accuracy across every topic or normal user prompt."},{"id":"GOOGLE26-01","evidenceClass":"operator_claim","claim":"Google reported that AI Mode had surpassed one billion monthly users by May 2026 and that its queries had more than doubled each quarter since launch.","population":"Google AI Mode users as defined by Google","period":"May 2026","source":"Google Search product announcement","url":"https://blog.google/products-and-platforms/products/search/search-io-2026/","caveat":"Operator-reported scale with no public user-count methodology in the announcement; it is not independent market share."},{"id":"GOOGLE25-01","evidenceClass":"product_capability","claim":"Google documents that AI Overviews and AI Mode can fan a question out into multiple related searches and supporting pages.","population":"Google Search AI features","period":"Documentation updated December 2025","source":"Google Search Central","url":"https://developers.google.com/search/docs/appearance/ai-features","caveat":"Documentation describes intended capability, not the frequency, ranking weight or completeness of fan-out."},{"id":"GOOGLE25-02","evidenceClass":"product_capability","claim":"Google says no special AI schema or machine-readable AI file is required; a page must be indexable and snippet-eligible, while standard SEO practices remain relevant.","population":"Site owners seeking visibility in Google AI features","period":"Documentation updated December 2025","source":"Google Search Central","url":"https://developers.google.com/search/docs/appearance/ai-features","caveat":"Eligibility does not guarantee crawling, indexing, selection or traffic."},{"id":"OPENAI26-01","evidenceClass":"product_capability","claim":"ChatGPT Search may automatically search, rewrite a request into targeted searches and display inline citations plus a Sources panel.","population":"ChatGPT Search users","period":"Help documentation viewed August 2026","source":"OpenAI Help Center","url":"https://help.openai.com/en/articles/9237897-chatgpt-search","caveat":"Search and citation behavior depends on the request, product state, location and plan limits."},{"id":"OPENAI26-02","evidenceClass":"product_capability","claim":"ChatGPT shopping can use intent, context, structured product metadata and merchant data to show products, comparisons and in some cases checkout.","population":"Eligible ChatGPT shopping users and merchants","period":"Help documentation viewed August 2026","source":"OpenAI Help Center","url":"https://help.openai.com/en/articles/11128490-shopping-with-chatgpt-search","caveat":"Availability varies; prices, stock, labels and generated summaries can be incomplete or stale."},{"id":"BING25-01","evidenceClass":"product_capability","claim":"Microsoft describes Copilot Search in Bing as a hybrid experience that combines synthesized answers, cited sources and conventional web results.","population":"Bing users with Copilot Search availability","period":"April 2025","source":"Microsoft Bing Search Blog","url":"https://blogs.bing.com/search/April-2025/Introducing-Copilot-Search-in-Bing","caveat":"Product description; it is not an independent quality or usage assessment."},{"id":"CLAUDE25-01","evidenceClass":"product_capability","claim":"Anthropic documents web search that retrieves current information and attaches direct citations to Claude responses.","population":"Claude users where web search is available","period":"2025-2026","source":"Anthropic","url":"https://www.anthropic.com/news/web-search","caveat":"Availability and limits vary by plan and location; citations do not independently prove answer accuracy."},{"id":"PERPLEXITY26-01","evidenceClass":"product_capability","claim":"Perplexity defines its answer engine around web retrieval, synthesized answers and linked citations, with deeper research modes for multi-source work.","population":"Perplexity users","period":"Help documentation viewed August 2026","source":"Perplexity Help Center","url":"https://www.perplexity.ai/help-center/en/articles/10352155-what-is-perplexity","caveat":"Self-description of a product, not a comparative accuracy result."},{"id":"ADOBE26-01","evidenceClass":"vendor_analytics","claim":"Adobe Analytics reported that generative-AI referrals to United States retail sites increased 693.4% year over year during the 2025 holiday season.","population":"United States retail traffic measured by Adobe Analytics","period":"November-December 2025 versus 2024","source":"Adobe Digital Insights","url":"https://business.adobe.com/blog/ai-driven-traffic-surges-across-industries","caveat":"Relative growth from an undisclosed small base; it is not AI's share of ecommerce traffic."},{"id":"ADOBE26-02","evidenceClass":"vendor_analytics","claim":"Adobe reported that referred retail visits from AI sources converted 31% more often and spent 45% more time on site than other traffic during the measured holiday period.","population":"United States retail traffic measured by Adobe Analytics","period":"2025 holiday season","source":"Adobe Digital Insights","url":"https://business.adobe.com/blog/ai-driven-traffic-surges-across-industries","caveat":"The channels are not randomly assigned; higher engagement can reflect different intent, users or merchant mix."},{"id":"V13-01","evidenceClass":"viton13_analysis","claim":"AI discovery is best measured with three separate ledgers: system capability, human behavior and publisher or merchant outcomes.","population":"Analytical framework for publishers and site owners","period":"2026","source":"VITON13 Research","url":"https://viton13.com/research/05-state-of-ai-search","caveat":"Editorial synthesis, not a population estimate."},{"id":"V13-02","evidenceClass":"viton13_analysis","claim":"The dominant 2026 pattern is addition and hybridization: AI answers are becoming a layer across search, chat, shopping and actions rather than a complete replacement for the open web.","population":"Cross-source interpretation of the report evidence","period":"Evidence through August 13, 2026","source":"VITON13 Research","url":"https://viton13.com/research/05-state-of-ai-search","caveat":"A time-bounded interpretation that may change as interfaces and measurement improve."}]}