Google’s search results are undergoing a big change. Instead of the familiar list of blue links, many users now see AI-generated summary answers – what Google calls “AI Overviews” – at the top of their search pages. Early data shows this shift is already draining clicks from websites: after AI Overviews rolled out, clicks to top-ranked sites dropped by over 34%. For the average internet user, it means getting instant answers without leaving Google. But behind that convenience lies growing unease about what this means for public access to information, the livelihoods of content creators, the future of SEO, and the balance of power over the web’s knowledge.
In this analysis, I explore the broader implications of Google’s AI Overviews. How does an AI-curated answer change the way people find information? Can journalists, publishers, and other creators sustain their work when fewer readers click through? What becomes of SEO and content discoverability in a world of “zero-click” searches? And ultimately, is it healthy for a handful of AI systems to gatekeep what information we all see? These questions cut to the core of whether Google’s AI is compatible with a vibrant, open web – or quietly undermining it.
From Ten Blue Links to One Answer
For two decades, Google Search served as a gateway to a plurality of sources – the blue links offering a selection of websites for any query. AI Overviews change that paradigm by condensing information from multiple sources into a single synthesized answer at the top of the page. Google touts this as a boon for users, claiming the AI summary helps people ask more complex questions and quickly find high-quality content. Indeed, not having to click through multiple sites to assemble an answer is convenient. Complex questions that once required scouring several pages can now be answered in one go, in plain language, right on Google’s screen.
However, this one-stop answer comes at the cost of breadth and context. When an AI Overview provides what looks like a definitive response, users may not feel the need to examine other sources or viewpoints. They might not even scroll down to see the traditional results anymore. This raises the stakes for accuracy – and there have been early missteps. In the beginning, Google’s AI Overview feature served up dubious or even dangerous misinformation in its responses, from suggesting people “add some glue” to pizza dough to advising they eat a small rock every day. Many users will accept an authoritative-sounding answer at face value, especially when it’s presented by Google. The public may be less aware of errors and less exposed to the nuance that multiple sources provide.
There’s also a concern about the diversity of information that users see. Google’s AI summaries do include citations, but they overwhelmingly draw on a narrow set of “trusted” websites. One analysis found that over 93% of links cited in AI Overviews come from the top 10 traditional search results. In theory this focus on authoritative sources helps reduce inaccuracies (Google has noted it only shows information backed by high-ranking results. But it also means the AI is largely amplifying the usual dominant voices. Niche blogs, emerging perspectives, or minority viewpoints that might have appeared lower on page one are less likely to surface in an AI-curated answer. Public access to information could become more homogenized – highly filtered through Google’s AI criteria – rather than the open exploration the web once promised.
CTR decline following introduction of AI Overviews (Ahrefs)
Content Creation in a No-Click World
The average click-through rate (CTR) for Google’s top organic search result plummeted from about 7.3% in March 2024 to 2.6% in March 2025 after AI Overviews rolled out. Fewer clicks mean fewer readers landing on the sites that actually produce the information – a worrisome trend for anyone who earns a living by publishing content online. In the past, capturing the #1 spot on Google could flood a site with traffic. Now, even being ranked first may yield only a trickle of visitors if the AI snippet already answers the query. This decline directly undermines Google’s earlier assurances that AI answers would increase engagement with web content; at least one Google executive had argued that the new AI results could boost click-through rates. That claim, as skeptics expected, has not borne out – if anything, the opposite is happening.
Multiple studies now confirm what many publishers feared: user clicks are dropping significantly when AI summaries appear. The recent analysis by Ahrefs (referenced above) found a 34.5% drop in click-through for the top organic result after AI Overviews launched. Ahrefs’ data – drawn from hundreds of thousands of searches – suggests this isn’t a minor blip but a broad pattern. Its content marketing director predicts click rates will only continue to slide as users grow more accustomed to getting answers straight from Google. And the top result isn’t the only casualty; when fewer people click the first link, the cascade effect means even fewer trickle down to the second, third, and beyond.
Google’s own interface often pushes the entire list of traditional results far down the page, sometimes requiring users to scroll past a full page of AI answer and related content before organic links appear. It’s no wonder that roughly 60% of searches now end without any click-through at all.
For publishers, journalists, and independent creators, this trend is sounding alarms. Their content is being scraped, summarized, and served up by Google, but fewer users are clicking through to give them the page views, ad impressions, or revenue opportunities needed to fund that content. If a site’s traffic drops by a third, its ad revenue and audience growth tend to drop in tandem. Some worry about a vicious cycle: as clicks and revenues fall, sites will have to cut back on content or put up paywalls, which in turn diminishes the freely accessible information that Google’s AI can draw upon. In extreme scenarios, AI-driven search could virtually eliminate traffic to many websites – one study found that when users get answers from a pure AI search engine (like OpenAI’s or Perplexity’s tools), referrals to news sites and blogs collapsed by 96% compared to classic Google search. Even a fraction of that impact in Google’s ecosystem would be devastating. To survive, content producers increasingly realize they might have to find new ways to align with the AI era of search.
Google AI Overview (Unite AI/Alex McFarland)
SEO in the Age of Generative Search
These changes also upend the traditional playbook of search engine optimization (SEO). For years, SEO experts optimized content to rank high on Google’s result page, believing that a top ranking guarantees attention. Now, even high-ranked pages can become invisible if they’re buried under an expansive AI answer box. Marketers are observing a dramatic rise in “zero-click” searches, with Google answering queries outright on the results page.
Furthermore, Google’s AI Overviews have been known to displace top-ranked links by as much as 1,500 pixels – about two full screens on desktop (and even more on mobile). In practice, this means a user might have to scroll a lot before they even see the first organic link. The fight for visibility has never been fiercer, and the old SEO formula has been thrown out the window.
Unsurprisingly, businesses and publishers are racing to adapt. The goal now is not just to rank, but to be featured inside the AI answer. We’re seeing the rise of what some call “Answer Engine Optimization” – tweaking content so that Google’s AI is more likely to pick it up in its summary. Google’s system heavily favors content it deems authoritative (nearly all sources cited by AI Overviews come from the first page of results, so the traditional pillars of SEO like quality content and backlinks remain important. But beyond that, content creators are experimenting with new tactics: writing in a Q&A format, adding structured data, and explicitly answering probable user queries within their text, hoping the AI will incorporate their material. In essence, site owners are learning to appeal to Google’s AI itself. The businesses that optimize now for AI-driven results will secure their position, while those that don’t risk becoming invisible.
Interestingly, if you do manage to be included as a source in an AI overview, the quality of traffic you receive might improve even as quantity falls. Early evidence suggests that visitors who click through from an AI-curated answer tend to be highly engaged. This makes sense – if the AI overview funnels a user to a specific article, it’s likely because that article squarely addresses what the user is looking for. In response to criticism that AI answers keep people on Google, the company has also started adding more direct links into the AI summaries and even integrating ads alongside them.
These changes are meant to drive users to publisher websites and reassure content creators that Google’s AI isn’t a dead-end for their traffic. Still, for most publishers, the balance of power has clearly shifted: it’s Google’s AI that orchestrates discoverability, and everyone is scrambling to work within (or around) that new reality.
AI Gatekeepers and the Web’s Future
The rise of AI-generated answers raises a profound question: who controls what information the public sees? In the era of blue links, Google’s influence was enormous but it still acted mostly as an index and referee – presenting a range of sources and letting users choose. With AI Overviews, Google (and its algorithms) take on a stronger editorial role, synthesizing and filtering knowledge before the user even clicks anything. If this model becomes the norm, the company’s AI effectively becomes the gatekeeper of the open web, and other AI search tools could play a similar role.
Google insists that “helping people discover content from the web remains central to our approach,” but the reality is that AI mediation gives it even more control over what gets shown, cited, or ignored. The major tech and AI providers are poised to become the primary gatekeepers of information access, setting the terms for which facts or opinions a searcher sees. For publishers and creators, reaching an audience may increasingly require appeasing those gatekeepers – conforming to the formats or guidelines that the AI favors.
This consolidation of informational power has ethical and economic ramifications. On the ethical side, there’s the issue of bias and diversity: if a handful of AI systems determine the “best” answer for millions of users, there’s a risk that minority views or emerging information get systematically filtered out. Media analysts warn that this trend could concentrate influence over global information and shrink the diversity of voices, essentially placing much of the world’s knowledge curation in the hands of a few U.S. tech companies.
On the economic side, AI-generated answers are built on content from the open web – content someone paid to create – yet those same answers may deprive content creators of the traffic and revenue they need to survive. It’s a paradox of value extraction: AI needs a thriving open web to feed it data, but by siphoning attention away from websites, it undermines the web’s very foundation. To further complicate matters, some AI firms have started striking deals with large publishers for direct access to their content. These licensing arrangements (often behind closed doors) could create a two-tier system where big, established outlets get compensated and prioritized in AI results, while smaller independent sites are left out in the cold.
Meanwhile, web scraping by AI models has skyrocketed, indicating that AIs are gobbling up more online content than ever, even as they send proportionally less traffic back. In other words, the AI is taking more and giving less.
What does all this mean for the future of the open web? There’s a real possibility that, if left unchecked, we head toward a web where quality information exists but is mostly consumed secondhand through AI intermediaries rather than directly on websites. The incentive to create could dwindle for many creators if their work rarely gets direct readership or reward. Over time, a shrinking pool of content might even degrade the quality of AI answers, since these models rely on fresh, diverse information to stay accurate. Avoiding that negative spiral will likely require conscious effort and new norms or policies.
Some suggest that Google and other AI search providers may need to share a portion of ad revenues with content publishers, or tweak their AI UIs to more prominently encourage exploration of sources. Regulators, too, are paying attention – drawing parallels to past battles over how news aggregators use publisher content. At the end of the day, preserving an open, vibrant web ecosystem in the age of AI will be a collective challenge.
Google’s AI Overviews represent an impressive leap in search convenience and capability, but ensuring that this innovation benefits users and content creators alike is the key. The web was built on an open exchange of information, and it’s in our best interest – as users, as creators, as a society – to make sure that exchange doesn’t become a one-way street controlled by AI. The coming years will determine whether Google’s AI-enhanced search can truly complement the open web, or whether it will slowly enclose it.
Google AI Overviews (FAQs)
1. How do AI Overviews differ from traditional search results?
They replace blue-link lists with AI-generated summaries, reducing the need to click through to websites.
2. What are the main benefits of using AI Overviews?
They offer fast, synthesized answers to complex questions directly in the search results.
3. How does AI Overviews impact SEO strategies?
They shift focus from ranking high to being cited in AI answers, making “Answer Engine Optimization” essential.
4. What types of questions are best suited for AI Overviews?
Multi-part, informational, or research-style questions that require synthesis across sources.
5. How does AI Overviews handle complex queries?
By drawing from multiple top-ranking sources to generate simplified, conversational responses.