Google now answers your search directly — so why bother clicking through to the actual websites it got the answer from?

This explores why people so often stop at the AI summary instead of clicking through to the sources behind it. The short answer from the corpus: the summary usually does its job. When the answer is already on the page, clicking feels unnecessary. The numbers agree with each other across very different methods. Pew's browsing-panel data shows people click a result 8% of the time when an AI summary appears, compared with 15% without one, and sessions are more likely to end with no click at all Do AI summaries on Google reduce clicks to actual websites?. A randomized field experiment found a 39.8% drop in clicks to publisher sites Do AI Overviews actually reduce clicks to publisher sites?. Ahrefs measured a 58% lower clickthrough rate for the top-ranked result How much are AI Overviews actually reducing organic search clicks?.

The first mechanism is attention. Eye-tracking work from RMIT and Microsoft shows the AI Overview now takes the 'golden triangle' at the top of the page, where eyes land first. The share of attention going to the first-ranked result falls from 31% to 9% Where do searchers look when AI Overviews appear?. The surprising detail is that people trusted the summary and the ranked links about equally. They weren't skipping sources because they distrusted them. The sources had simply been pushed out of view. Even Wikipedia, which often supplies the facts these summaries repeat, lost about 5% of its search traffic to what researchers call 'answer-layer intermediation': the question gets satisfied before anyone reaches the original Does AI search summaries divert traffic away from Wikipedia?.

The most counterintuitive point comes from outside search entirely. Nextdoor found that its better, more informative AI summaries produced *fewer* clicks, because a good summary removes the reason to open the full item Does better summary writing actually increase user engagement?. So low clickthrough isn't a sign that AI Overviews are failing. It can be a sign that they're working, and that improving them will cut clicks further. The field experiment adds a twist: clicks fell, but people didn't rate their searches as any better Do AI Overviews actually reduce clicks to publisher sites?. Users lose the trip to the source without a clear gain in return.

It also matters what's lost when people don't click. Seven experiments with more than 10,000 participants found that people who learned from LLM summaries came away with shallower knowledge and gave sparser advice than people who searched the web themselves Does learning from AI summaries produce shallower knowledge than web search?. The effort of clicking, reading and comparing sources seems to be part of how understanding forms. Interface design may help. In one experiment, a persistent sidebar of sources improved critical thinking as more citations were added, while compact hover cards made it harder for people to combine information from several sources Can interface design reverse citation overload's harm to critical thinking?. How sources are displayed affects whether anyone uses them.

The effects go beyond individual users. Publishers expect search referrals to fall 40–43% over three years Will AI Overviews reduce search referral traffic to publishers?. A Munich court has ruled that AI Overview statements are Google's own content, not neutral pointers to someone else's Can courts hold Google liable for AI Overview false claims?. That ruling names the underlying shift: once the search page answers the question instead of pointing to other sites, the sources become background material, and both the traffic and the legal responsibility move to the company writing the answer.