From Google to ChatGPT: Students are changing how they search for knowledge

Students worldwide are changing how they find information, with artificial intelligence tools such as ChatGPT increasingly becoming part of the search process

A new study led by researchers at Bar-Ilan University suggests that the change is particularly noticeable when students look for explanations of physics concepts.

The study, published in Physical Review Physics Education Research, examined search activity across more than 20 countries between 2022 and 2025.

The research found a decline in Google searches for physics topics, along with a similar decline in visits to related Wikipedia pages.

A change from search to conversation

Researchers say the change reflects how students seek knowledge, not necessarily that they are learning less.

Traditional online research often requires students to search for information across several websites, compare sources and piece together an explanation. AI tools offer a different experience by providing a direct, conversational response to a question.

This could make finding information faster and more accessible, particularly for students who may struggle to identify reliable educational resources online.

Bigger changes outside English-speaking countries

The research showed a difference between English-speaking and non-English-speaking countries.

In countries such as the United States, the United Kingdom and Australia, Google searches for many physics subjects remained comparatively stable. In several non-English-speaking countries, however, searches declined much more significantly.

Language may be an important factor. A lot of the educational material available online has traditionally been produced in English, meaning students who speak other languages can face an additional barrier when trying to understand complex subjects.

Generative AI could help reduce that barrier by providing explanations in a student’s preferred language. This may give learners greater access to scientific information without relying as heavily on English-language websites.

Not every subject is moving to AI

The study also found that the decline in Google searches varied by physics topic.

Searches fell more sharply for subjects that can largely be explained through text and logical reasoning. Topics that depend heavily on diagrams, graphs and other visual information showed smaller declines.

The researchers believe this may reflect AI systems’ current strengths. Large language models are particularly effective at producing written explanations, but tasks involving complex visual information can still be more challenging.

What does this mean for education?

The findings offer an early look at how generative AI may be changing students’ everyday learning habits.

The research team used Google Trends to measure changes in search activity and examined Wikipedia page views in seven languages to see whether the pattern extended beyond Google. The similar trends on Wikipedia suggest the change may reflect a wider shift in how students access information.

As AI becomes increasingly integrated into education, teachers, universities and educational publishers may need to reconsider how learning resources are designed. Instead of simply directing students toward information, future educational approaches may need to account for a world in which students can ask an AI system to explain, simplify or translate complex ideas instantly.