logo
Web traffic collapses as AI steps in: Google is changing how we click

Web traffic collapses as AI steps in: Google is changing how we click

Google's push into AI-powered search is reshaping the internet—and not in the way publishers or web creators might have hoped.
A new study by the Pew Research Centre, which examined the online behaviour of 900 US adults, reveals that Google's AI Overviews, the automatically generated summaries now appearing in many search results, are drastically reducing the number of users who click through to traditional websites.
That means less traffic for publishers, fewer pageviews for advertisers, and a growing concern over who benefits from AI-enabled search.
Do users click less on links when AI summaries appear?
In March 2025, only 8 per cent of searches featuring an AI Overview resulted in a link click, Pew found. That's nearly half the rate of searches without the feature. And even when AI summaries do include citations, users rarely interact with them—just 1 per cent clicked on any embedded link.
The data points to a fundamental shift in user behaviour. Google's AI is not just organising information—it's becoming the destination.
'This is a classic case of convenience cannibalising discovery,' said one digital strategy analyst quoted in the Pew study. 'If users get what they need in a sentence or two, there's no reason to click further—and that's a revenue black hole for anyone not named Google.'
Is AI replacing traditional website visits?
Beyond clicks, Pew's research found that 26 per cent of users who encountered an 'AI Overview' ended their browsing session immediately, compared to 16 per cent when no AI content appeared.
The implication: AI is not just answering questions; it is shutting down curiosity. The findings also suggest that AI-generated content may be doing more than simply guiding users. It may be replacing the need to visit external websites altogether.
For content creators, the concern is not just about fewer eyes—it is about fewer opportunities to engage, explain, and build trust.
Which sources are cited most by Google's AI Overviews?
Pew also analysed the types of sources most frequently cited in AI Overviews. Wikipedia, YouTube, and Reddit collectively accounted for 15 per cent of citations, with Wikipedia appearing more frequently in AI content than in standard search results (where the three sites combined made up 17 per cent).
Government websites appeared three times more often than in traditional search results.
News outlets, meanwhile, saw no increase, holding steady at 5 per cent.
That uneven distribution raises questions about source diversity and editorial bias. AI Overviews appear to be prioritising broadly accepted or authoritative sources, but potentially at the cost of lesser-known publishers.
When does Google trigger an AI Overview in search?
Interestingly, AI Overviews are not triggered across the board. Only 18 per cent of all searches in the study included an AI summary. The more specific the query, the more likely the AI feature activates. Searches with ten or more words triggered AI summaries 53 per cent of the time, compared to just 8 per cent for short queries.
Full, question-style queries—such as those beginning with 'what' or 'why'—were also more likely to prompt an AI response.
On average, AI Overviews were around 67 words long, though they ranged anywhere from seven words to a more detailed 369-word response.
Will AI summaries hurt publishers and educators long term?
AI Overviews offer convenience and speed for users. While there is no suggestion that users are learning less, it does seem they are, increasingly, learning differently—and from fewer sources. That shift could have long-term consequences for how knowledge is produced, distributed and monetised.
For the wider internet ecosystem, however, this has raised alarms—especially newsrooms, educators, and independent websites. Reduced traffic means reduced visibility and, for many, declining revenue.
Google is making search faster, but it may also be undermining the very ecosystem that makes the web useful.
Orange background

Try Our AI Features

Explore what Daily8 AI can do for you:

Comments

No comments yet...

Related Articles

Dominance of Amazon and Microsoft in cloud harming competition, UK says
Dominance of Amazon and Microsoft in cloud harming competition, UK says

The Hindu

time20 minutes ago

  • The Hindu

Dominance of Amazon and Microsoft in cloud harming competition, UK says

The dominant position of Amazon and Microsoft in cloud computing is harming competition, with their impact exacerbated by technical and commercial barriers to switching, an inquiry group from Britain's antitrust regulator said. The Competition and Markets Authority (CMA) group said on Thursday the regulator should investigate whether to designate the two with strategic market status (SMS) in cloud services, which would give it new powers to intervene. It noted, however, that the CMA has said it will not consider new SMS investigations, which are conducted by its Digital Markets Unit (DMU), until early next year. Microsoft was singled out in its final report for licensing practices that the panel said adversely impacted Amazon Web Services (AWS) and Google. The group said in January that Microsoft was using its dominance in enterprise software, such as Windows Server and Microsoft 365, to limit competition by charging licensing fees when its services were used on rival platforms. Microsoft and AWS have 30-40% market shares in cloud services such as processing, storage and networking, it said. Google is the third main provider, but it has a smaller share of 5-10%. "Measures aimed at Microsoft and AWS would address market-wide concerns," the CMA group said. The cloud computing industry has been scrutinised by regulators on both sides of the Atlantic. In Europe, Microsoft clinched a 20-million-euro deal last year to settle a complaint about its licensing practices, averting an antitrust investigation and potential hefty fine. The company said the CMA group's report "misses the mark again, ignoring that the cloud market has never been so dynamic and competitive, with record investment, and rapid, AI-driven changes". "Its recommendations fail to cover Google, one of the fastest-growing cloud market participants," a spokesperson said. Amazon said "clear evidence of robust competition" had been disregarded. "The action proposed by the Inquiry Group is unwarranted and undermines the substantial investment and innovation that have already benefited hundreds of thousands of UK businesses," a spokesperson said. But it noted the group had recognised that action needed to be taken over Microsoft's licensing practices. Google said the conclusive finding that restrictive licensing harmed customers and competition was a "watershed moment". "Swift action from the DMU is essential to ensure British businesses pay a fair price and to unleash choice, innovation and economic growth in the UK," said Chris Lindsay, Google Cloud's vice president for customer engineering EMEA. The Open Cloud Coalition and the Coalition for Fair Software Licensing said the CMA should take action quickly. "Given the alarming anticompetitive behaviour it has identified, the current plan to start this process in early 2026 is nowhere near sufficient," said Nicky Stewart, senior advisor to the Open Cloud Coalition.

Microsoft stops 30-year-old practice and drops Apple, IBM, Meta, Nvidia from this its list of ...
Microsoft stops 30-year-old practice and drops Apple, IBM, Meta, Nvidia from this its list of ...

Time of India

time28 minutes ago

  • Time of India

Microsoft stops 30-year-old practice and drops Apple, IBM, Meta, Nvidia from this its list of ...

Microsoft has opted not to name its rivals in its latest annual report, in a significant shift from a nearly 30-year tradition, a report has pointed out. The 101-page document, released this week, contains no direct references to competitors like Apple, Google and IBM, nor newer private challengers such as Anthropic or Databricks. This marks a notable departure for the 50-year-old technology giant, which has listed its competitors in annual reports since at least 1994. Last year's report, for instance, explicitly designated over 25 companies as rivals. What is the 'Big change' that Microsoft has brought in its annual report Instead of specific names, Microsoft's latest report broadly states that it 'faces competition in a wide variety of markets,' including productivity software, PC operating systems and cloud infrastructure. A Microsoft spokesperson informed CNBC that this updated format aims to address larger categories. 'The new format reflects the fast-moving nature of the markets in which the company operates', CNBC cited a company spokesperson as saying. Previously, these specific disclosures offered insights into how companies positioned themselves against the software titan. In a notable shift just last year, Microsoft even began referring to key partner OpenAI as a competitor after the AI startup launched a web search feature. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Moose Approaches Girl At Bus Stop In Chittagong - Watch What Happens Happy in Shape Undo Despite this change in official documentation, Microsoft executives continue to track and comment on rival developments. CEO Satya Nadella referenced Amazon during Microsoft's earnings call. Similarly, Scott Guthrie, executive vice president of Microsoft's cloud and AI group, referred to Nvidia's powerful GB200 NVL72 systems. Tech giants that name and do not specifiy their competitors The this practice of not naming isn't universally followed. For example, according to CNBC, Microsoft's omission breaks from a common industry practice still observed by peers like Apple, Meta and Nvidia, all of whom continue to name companies they compete against. Meanwhile, Amazon hasn't identified competitors in its annual reports since 1999, Tesla stopped in 2020, and Alphabet (Google's parent) ceased after 2022. The new approach to competitive disclosure comes as Microsoft's shares climbed, pushing the company's market capitalisation past the $4 trillion mark. AI Masterclass for Students. Upskill Young Ones Today!– Join Now

OpenAI kills ChatGPT feature that exposed personal chats on Google: All you need to know
OpenAI kills ChatGPT feature that exposed personal chats on Google: All you need to know

Mint

time38 minutes ago

  • Mint

OpenAI kills ChatGPT feature that exposed personal chats on Google: All you need to know

OpenAI has just announced that it is killing a new feature that made ChatGPT conversations available on Google Search. The feature was rolled out quietly earlier in the year but caused much uproar recently as many private conversations started showing up on Google search results, putting the privacy and safety of users at risk. OpenAI Chief Information Security Officer (CISO) Dane Stuckey, in a post on X (formerly Twitter), informed users about taking down the feature from the popular chatbot. 'We just removed a feature from @ChatGPTapp that allowed users to make their conversations discoverable by search engines, such as Google. This was a short-lived experiment to help people discover useful conversations. This feature required users to opt in, first by picking a chat to share, then by clicking a checkbox for it to be shared with search engines (see below),' Stuckey wrote. 'Ultimately we think this feature introduced too many opportunities for folks to accidentally share things they didn't intend to, so we're removing the option,' he added. Stuckey also noted that OpenAI is working with search engines to remove the already indexed content from the web. OpenAI had rolled out an update to ChatGPT that gave users the option to make their chats discoverable on search engines. While the company has clarified that this was an opt-in feature and not enabled by default, it seems that many users inadvertently clicked on this option while sharing their chats with friends or family. As OpenAI CEO Sam Altman rightly pointed out in a recent podcast, 'people talk about the most personal shit in their lives to ChatGPT.' A new report by Fast Company revealed that there were over 4,500 conversations indexed on Google. While many of them did not contain personal data, others did — including identifiable details like names, places and more — that could put users at risk. Moreover, even if a user deleted the conversation link or deleted the conversation altogether, it would not guarantee that it was no longer indexed by Google. In fact, the conversation would likely remain public until Google updates its index.

DOWNLOAD THE APP

Get Started Now: Download the App

Ready to dive into a world of global content with local flavor? Download Daily8 app today from your preferred app store and start exploring.
app-storeplay-store