Ask Google a question in AI Mode and you do not get a results page at all. The query is split into a fan of sub-searches, Gemini reads across the sources, and a single synthesised answer comes back inside a chat thread you can keep interrogating. That is the distinction that matters to a publisher. AI Overviews still sit on top of ten blue links a reader might click; AI Mode replaces the links page entirely with a conversation. Your content is read, summarised and occasionally cited, but the page itself is no longer where the user ends up - which means the page is no longer where you can reliably reach them.
What is Google AI Mode?
Google AI Mode is a conversational search experience, powered by the Gemini model, that handles a query as a dialogue rather than a list of results. Instead of typing keywords and scanning links, the user asks a full question, gets a synthesised answer, and asks follow-ups in the same thread. Under the surface Google uses a "query fan-out" technique: it breaks the question into multiple related sub-queries, runs them in parallel, reads across the results, and assembles one response with a handful of supporting links. AI Mode is multimodal, so a user can search with text, voice or an image, and it is built to handle longer, more complex questions than a single search box was ever designed for. Google rolled it out to everyone in the United States in May 2025 without a Labs sign-up, and has since expanded it internationally and made it a standing part of the Search experience.
How is AI Mode different from AI Overviews?
AI Overviews and AI Mode are two different surfaces, and the difference decides how much traffic a publisher loses. An AI Overview is the AI-generated summary box that appears at the top of an otherwise normal results page, so the blue links the reader might click are still present underneath it. AI Mode is a separate, fully conversational experience that replaces the results page with a chat thread, so there is no list of links to fall back to. They do not even draw on the same sources consistently: one 2026 analysis found that AI Mode and AI Overviews cite the same URLs only about 13.7 per cent of the time, despite reaching broadly similar conclusions in roughly 86 per cent of cases. For a publisher that means visibility in one surface does not guarantee visibility in the other, and the surface that contains the user most tightly - AI Mode - is also the one that gives them the least reason to leave.
How big is AI Mode, and how fast is it growing?
AI Mode reached one billion monthly users within roughly a year of its public launch, according to figures Google shared in May 2026, with query volume reported to be more than doubling each quarter since it shipped. For comparison, AI Overviews reaches around 2.5 billion users a month. So AI Mode is not yet the larger surface, but it is the faster-growing and the more contained one, and its trajectory points the same way as every other AI search product: towards answers that resolve inside Google rather than journeys that send the user out to a publisher. The strategic read for media businesses is that this is not a fringe beta to monitor. It is already one of the largest search surfaces on the web and it is being normalised as part of how a billion people search.
What does AI Mode do to publisher traffic?
AI Mode reduces referral traffic by resolving the query before the user ever needs to click. When the answer, the follow-ups and the next step all happen inside the thread, the visit that used to carry an ad impression or an affiliate link simply does not occur. Direct measurement of AI Mode click-through is still thin, but the closest proxy is alarming: the Pew Research Center, tracking 68,000 real searches, found users clicked a result only 8 per cent of the time when an AI summary was present, against 15 per cent when it was not, and clicked a source link inside the summary just 1 per cent of the time. The wider trend is the same. Ahrefs measured a 58 per cent fall in click-through on the top-ranking page for queries carrying an AI Overview in February 2026, and Chartbeat reported roughly a third fewer search referrals to publishers worldwide. AI Mode, with no links page at all, sits at the sharper end of that curve.
Is Google putting ads in AI Mode?
Yes - Google has confirmed it is bringing ads into AI Mode, which is the development publishers should watch most closely. Google announced in May 2025 that it would test advertising in AI Mode, and has said advertisers already running Performance Max, Shopping and Search campaigns on broad match are eligible to appear there, with US users seeing Search and Shopping ads woven into AI Mode responses across desktop and mobile. The significance for publishers is structural, not incidental. Google is building the ability to monetise the answer surface itself - the conversational reply assembled from publishers' content - while the publishers whose work feeds that answer receive neither the click nor a share of the ad that runs beside it. The value is being captured at the point of the answer. The open question for the rest of the market is who else gets to monetise that moment, and where.
How can publishers respond to AI Mode?
The practical response runs along three lines. First, measure server-side. AI surfaces strip or suppress the referrer, so the reliable record of who is reading you lives in your logs and at the CDN edge, not in client-side analytics that increasingly file this traffic as "direct". Second, structure content for retrieval rather than for ranking alone: lead each section with a direct, self-contained answer, use question-format headings, add Article and FAQ structured data, and include original data points an AI system can cite you for. This improves the odds of being the named source inside an AI Mode answer even when there is no click. Third, treat the read itself as the monetisable event. If the user resolves their question inside Google's surface and never lands on the page, monetisation has to move to where the content is actually consumed. This is the layer blankspace works in: detecting AI and agent traffic at the CDN edge and placing contextual brand mentions into the answer, so value is captured at the moment of retrieval rather than lost when the visit never happens.
Frequently asked questions
What is the difference between Google AI Mode and Gemini?
Gemini is Google's underlying AI model and its standalone assistant app; AI Mode is the search experience inside Google Search that uses Gemini to answer queries conversationally. Put simply, Gemini is the engine and AI Mode is one of the places that engine is applied. A publisher cares about AI Mode specifically because it sits inside Search, where the audience that used to arrive as referral traffic now begins and often ends its query.
Is AI Mode replacing the normal Google results page?
Not entirely, but it is being positioned as a primary way to search rather than a side feature. The classic results page with blue links still exists, and AI Overviews still appear on top of it, but AI Mode offers a separate conversational path that resolves the whole query inside Google. As adoption grows, a rising share of searches never reach a list of links at all, which is why publishers are seeing referral declines even where their content is being read.
Does AI Mode cite and link to publisher sites?
AI Mode does include supporting links and citations alongside its answers, but they are surfaced selectively and clicked rarely. Pew Research found source links inside AI summaries were clicked only about 1 per cent of the time, and AI Mode draws on a different set of URLs from AI Overviews in most cases. So being cited in AI Mode is valuable for brand visibility, but it rarely delivers the click volume that the old results page did.
How is AI Mode different from ChatGPT or Perplexity?
AI Mode is Google's conversational answer built directly into Search and fed by Google's index, whereas ChatGPT and Perplexity are separate assistants users go to deliberately. The publisher impact is similar across all of them - the content is read to build an answer and the human often never visits the page - but AI Mode matters most by sheer scale, because it sits inside the search habit of a billion-plus people rather than requiring them to switch to a different product.
How can publishers prepare for AI Mode now?
Start by reading server and edge logs rather than relying on client-side analytics, so you can see the AI traffic that conventional tools miss. Then structure your content for extraction, with direct-answer openings, question-format headings and structured data, to improve the chance of being the cited source. Finally, plan for monetisation that does not depend on a click, because a growing share of your audience now gets what it needs inside the answer. Edge-level approaches that monetise the AI response itself, such as blankspace, are designed for exactly this shift.