Numbers, Facts and Trends Shaping Your World

Campaigns » The Briefing » Study: AI assistants often misrepresent news content

Study: AI assistants often misrepresent news content

☀️ Happy Thursday! The Briefing is your guide to the world of news and information. Sign up here!

In todays email:

  • Featured story: New research shows AI assistants often misrepresent news content
  • In other news: Trump refiles defamation suit against The New York Times
  • Looking ahead: The New York Times to debut vertical video in-app “Watch” tab
  • Chart of the week: Younger Americans more likely to say they see inaccurate news from AI chatbots

🔥 Featured story

New research from the European Broadcasting Union and the BBC finds that AI assistants often misrepresent news content. The study analyzed 3,000 responses to questions about the news from AI assistants, including ChatGPT, Copilot, Gemini and Perplexity. According to the research, 81% of the AI responses had some kind of problem, including 45% that had at least one significant issue.

Americans who get news from AI chatbots have mixed experiences with the news they get there – particularly when it comes to perceptions of its quality, according to a recent Pew Research Center survey. About half of U.S. adults who get news from AI chatbots say they at least sometimes come across news there that they think is inaccurate. And a third of those who use chatbots for news say they generally find it difficult to determine what is true and what is not, while fewer (24%) say they find it easy. The largest share (42%) say they aren’t sure.

📌 In other news

📅 Looking ahead

The New York Times announced that starting this week, its main app will host a stand-alone section for short-form vertical videos. The “Watch” tab will feature videos generally shorter than three minutes that span various areas of the Times’ coverage, including news, opinion, lifestyle and sports. It comes as part of a larger plan for the outlet to reach 15 million subscribers by 2027.

A growing share of Americans get news from social media platforms that feature video, such as TikTok, Instagram and YouTube, according to an August Pew Research Center survey. For example, one-in-five U.S. adults now say they regularly get news from TikTok, up from just 3% in 2020.

📊 Chart of the week

This week’s chart looks more closely at Americans’ perceptions of the accuracy of news they get from AI chatbots. A recent Center survey found that younger adults are more likely to say they see inaccurate news there. Among Americans who get news from AI chatbots, 59% of adults under 30 say they at least sometimes come across news they think is inaccurate, compared with 36% of those 65 and older.

Unlike with age, there are no major differences on this question by political party or education level.

👋 That’s all for this week. 

The Briefing is compiled by Pew Research Center staff, including Naomi Forman-Katz, Jacob Liedke, Christopher St. Aubin, Emily Tomasik, Joanne Haner, and Sawyer Reed. It is edited by Michael Lipka and copy edited by David Kent.

Do you like this newsletter? Email us at journalism@pewresearch.org or fill out this two-question survey to tell us what you think.

Icon for promotion number 1

Sign up for The Briefing

Weekly updates on the world of news & information