AI is becoming part of how people look for information, but the evidence does not show that most people now ask a standalone chatbot before they ever use Google. Many people report using chatbots, and many encounter AI summaries within search itself. Those are different routes—and the available studies measure adoption, exposure and clicks more clearly than they measure which route people choose first.
“Ask AI” can mean two different things
Someone may open ChatGPT, Gemini or Copilot and type a question. Or they may search in Google and see an AI-generated answer at the top of the results. In the second case, they have encountered AI without choosing a separate chatbot or necessarily changing how they start a search.
As an Amazon Associate I earn from qualifying purchases.
That distinction matters when interpreting claims about people asking AI before Google. Surveys can tell us whether people say they have used chatbots or seen AI summaries; browsing panels can record what people do after a search results page appears. Neither, by itself, establishes the order in which a person uses different tools across all their information searches.
What people report using—and seeing
Recent U.S. surveys point to broad use of AI for information, but their measures are not interchangeable:
#1 Best Overall
| Measure | Finding | What it tells us |
|---|---|---|
| Ever used an AI chatbot | 49% of U.S. adults, in a Pew Research Center survey of 5,119 adults conducted February 17–23, 2026. | Reported chatbot adoption; not when respondents use one relative to Google. |
| Use chatbots to search for information | 42% of U.S. adults, in the same Pew survey. | Reported use for an information-seeking purpose; not proof that the chatbot is the first stop. |
| Read AI summaries at the top of search results | 60% of U.S. adults, in the same Pew survey. | Exposure to AI within search, not standalone chatbot use. |
| Use AI to find information at least sometimes | 60% of U.S. adults and 74% of adults under 30, according to AP-NORC polling in 2025. | A broad self-reported measure that does not specify whether AI comes before Google. |
Sources: Pew Research Center, 2026; AP-NORC, 2025.
The clearest conclusion is that AI has become a familiar information route for many people. These figures do not show that chatbot use has displaced Google or that most users consistently begin with a chatbot.
What happens after Google shows an AI summary?
A Pew Research Center analysis offers a more direct behavioral measure. It examined browsing by 900 U.S. adults across 68,879 unique Google searches in March 2025. Traditional result links were clicked in 8% of visits to Google pages with an AI summary, compared with 15% of visits to pages without one. Links inside the AI summary were clicked in 1% of visits to pages that included a summary.
These are observed next-page actions in that panel, not results for every Google user. The study does not establish why someone did not click a link: they may have found what they needed, continued browsing elsewhere, refined the query or stopped. The difference in click rates also does not, on its own, prove that AI summaries caused the change.
Free tools Windows power users keep installed
One-click scans. No signup required.
Rank #2
Pew Research Center’s account of the browsing analysis describes the panel’s behavior and its limits.
AI answers are visible even when people do not open a chatbot
In a 2025 survey across six countries, 54% of respondents said they had seen an AI-generated answer in response to a search query in the previous week. Reported exposure ranged from 70% in Argentina to 64% in the UK, 61% in the U.S. and 29% in France. The Reuters Institute noted that Google’s AI Overview rollout status during fieldwork was one factor behind the lower French figure.
In the same report, weekly exposure to AI-generated search answers was higher than reported weekly use of the standalone AI tools covered by the survey, at 34%. The Institute also cautioned that some respondents might not distinguish generated summaries from conventional search results. Together, these points help explain why the shift can feel widespread even when a person has not deliberately adopted a chatbot: AI may arrive inside a product they already use.
Exposure depends on country, rollout and the period measured, so these results should not be treated as a universal or permanent rate. The Reuters Institute’s 2025 report gives the survey context.
Recommended Free Tools
Do people use ChatGPT instead of Google?
There is no single answer in the figures above. A chatbot and a search engine offer different starting points: a chatbot is opened intentionally for a conversational response, while a search engine presents results and may also include a generated summary. The studies here do not provide a broad, task-matched comparison of accuracy or trustworthiness, nor do they show which tool is better for every kind of question.
The best route depends on what the reader needs. A synthesized response can be a convenient starting point; a results page is useful when the goal is to find a specific site or compare sources. For consequential claims, inspect the cited material rather than treating a fluent answer as its own evidence. That is practical advice, not a finding that these studies tested.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What the shift may mean for publishers
Fewer clicks on Google results pages that display AI summaries are consistent with publishers’ concern that some readers may get information without visiting source sites. Separately, the Reuters Institute’s 2026 trends report cites Chartbeat data from more than 2,500 sites showing Google organic-search traffic fell 33% globally and 38% in the U.S. between November 2024 and November 2025.
The Reuters Institute says it is unclear how much of that decline is due to AI Overviews. The referral figures describe a change over time, not a measured causal share attributable to AI; other changes in search and referral behavior may contribute. The 2026 report sets out the trend and that qualification.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsGoogle’s explanation—and what it does not establish
Google says its AI features encourage more involved queries. In an August 2025 post, the company wrote: “With AI Overviews and more recently AI Mode, people are able to ask questions they could never ask before.” Google also said its internal data show people are happier with the experience and search more than ever. Those are Google’s own claims; the post does not publish the underlying measurement details in that passage, so they should not be mistaken for independent evidence that AI has made people abandon or prefer one search route.
Google’s post on AI in Search presents the company’s view of the feature and its performance.
Quick Recap
How to read the “AI-first” trend
- Separate use from exposure: opening a chatbot is not the same as seeing a generated answer inside a search engine.
- Separate what people say from what they do: surveys capture reported use or recall; browsing panels capture navigation, not every motivation behind it.
- Keep the scope attached to every statistic: the cited results concern particular populations, countries and study periods, not all users everywhere.
- Do not equate no click with satisfaction: a person’s next action does not reveal whether an answer was sufficient or whether they left to continue elsewhere.
- Check source links for important claims: a generated summary can orient a search, but following sources lets readers examine evidence and context themselves.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
