DigestAI news desk

Cut through the AI noise.

Marketing & Small Business9 min read

Study finds AI summaries reduce user knowledge and critical thinking

A 2025 study by Wharton professors Shiri Melumad and Jin Ho Yun, published in PNAS Nexus, indicates that users relying on AI summaries retain less information and produce lower-quality advice than those using standard search results. The research involved 10,462 participants across seven experiments. Even when the factual content was identical, participants using AI summaries spent less time…

1 source

Key points

  • Wharton study of 10,462 people found AI summary users knew less and wrote sparser advice than search users.
  • Pew Research found AI summaries reduced clicks on organic results to 8% from 15% in March 2025.
  • Users ended browsing sessions 26% of the time when AI summaries appeared, compared to 16% without them.

These findings align with data from Pew Research Center, which tracked 900 U.S. adults in March 2025. Pew found that when AI summaries appeared in Google search results, users clicked on normal organic results only 8% of the time, compared to 15% without summaries. Additionally, 26% of browsing sessions ended immediately after viewing an AI summary, versus 16% without one. Pew notes this is an association, not proven causation, and covers a single month of U.S. Google usage.

The article argues that this shift removes the "friction" of traditional research, which previously provided signals about the depth and reliability of information. For marketers, this means inbound leads may be "confidently underinformed," having skipped the educational content that used to guide them. The author suggests that content strategy must adapt, as the traditional funnel of beginner-to-advanced content no longer matches the user's journey, which now begins with an AI-generated summary.

Full story from Search Engine Journal · by Duane ForresterOpen source ↗

LLMs Are Time Machines That Don’t Tell You How Far You Went via @sejournal, @DuaneForrester

Search Engine Journal · 24 September 2026

Loading the full article…

This text was published by Search Engine Journal and written by Duane Forrester. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗

Topics · follow one to build your own front page
WhartonPew Research CenterGoogleMicrosoft ResearchCarnegie MellonYaleShiri MelumadJin Ho YunDirk LewandowskiDuane Forrester

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.

Comments

via GitHub Discussions

More in Marketing & Small Business

All →

Related stories