AI Search

AI search refers to information discovery systems powered by large language models that generate synthesized, contextual answers rather than traditional lists of links. Products like Perplexity, ChatGPT Search, Google's AI Overviews and AI Mode, and Claude represent a fundamental shift in how people find and consume information — one that is rapidly evolving into a broader category of AI answer engines.

The difference isn't just cosmetic. Traditional search engines return ranked lists of web pages, requiring the user to click through and synthesize information themselves. AI search systems read, analyze, and synthesize information on the user's behalf, delivering answers that may draw from dozens of sources—often without the user ever visiting a website. This is the mechanism behind zero-click search. The effect on clicks is measurable where it has been studied: Ahrefs (February 2026; 300,000 keywords) found that the presence of an AI Overview is associated with a 58% lower click-through rate for the top organic result, up from 34.5% in its April 2025 study.

The Answer Engine Era

By 2026, AI search has evolved from a novelty into a routine information-seeking behavior. Pew Research (June 2026) found that 49% of US adults use AI chatbots and 24% use them daily, and that 60% read AI summaries in search results. Perplexity pioneered the answer-engine format — providing sourced, synthesized responses with inline citations. ChatGPT Search integrated web browsing directly into conversational AI, letting users ask follow-up questions and drill deeper. Google AI Overviews placed AI-generated summaries above traditional search results. They do not appear everywhere: Semrush measured them on 15.69% of keywords in November 2025, down from a peak of 24.61% that July, but Seer Interactive (2026; 5.47 million queries) found them on 85.9% of question-format queries and 95.4% of comparison queries, against 5% of transactional ones. Google AI Mode, a fully conversational search surface, has scaled fastest. Google said at I/O in May 2026 that "AI Mode has surpassed one billion monthly users, with queries more than doubling every quarter," and since January 2026 the "Show more" control on an AI Overview opens AI Mode directly. Claude and other AI assistants serve as personal research tools that synthesize complex topics on demand.

Fan-Out and the Parallel Discovery Universe

AI search engines do not simply summarize the results for the query a user typed. Through query fan-out, the engine splits the question into several related sub-queries, retrieves results for each, and cites the pages that recur across them. The consequence is that the cited sources drift away from the familiar first page of results. Ahrefs (March 2026; 863,000 keyword results pages) found that 37.9% of the URLs cited in Google AI Overviews rank in the organic top 10 for the original query, 31.2% rank between 11 and 100, and 31.0% rank beyond position 100. The top-10 share had been roughly 76% in July 2025. BrightEdge (February 2026), with a different method, puts the overlap near 17%. For brands and creators, this means being visible in AI search requires more than ranking for a head term—giving rise to Generative Engine Optimization (GEO).

The value of the traffic that remains is less settled than its volume. No current primary source supports a universal conversion multiplier for AI search visitors, and published figures should be treated with caution. What has been measured is narrower: Seer Interactive found that a page cited inside an AI Overview earns 120% more organic clicks per impression than one that is not cited, and Similarweb (June 2026; US desktop, consumer sectors) reported that people shown a brand in a ChatGPT recommendation were 2.5 times more likely to visit its site within seven days, with 55.9% of those visits arriving through branded search rather than a link in the answer. Much of AI search's influence therefore shows up indirectly, and direct AI referral traffic understates it.

What Each Engine Cites

The engines draw on noticeably different sources. Ahrefs publishes the share each domain holds among the 50 most-cited sources per engine; its September 2026 editions show community and video platforms leading everywhere, in different proportions.

These shares are volatile and should be read with their dates attached. YouTube matters far more to Google's surfaces than to ChatGPT, Reddit is prominent on all three, and Wikipedia remains a steady reference source. Which pages get cited, and how much each one shapes the answer, is covered under AI citations.

From Search to Action

As AI search evolves into the agentic web, the systems don't just recommend—they act. An AI agent can evaluate options, compare prices, execute transactions, and manage follow-up, collapsing the entire funnel from discovery to purchase into a single interaction. The boundary between search engine, shopping assistant, research analyst, and personal agent is dissolving. This convergence represents the most significant shift in internet discovery since Google displaced directories in the early 2000s, and brands that fail to optimize for AI-mediated discovery risk becoming invisible to the next generation of consumers.