MJC director Benjamin Toff presented the findings at the second Summit on AI, Ethics and Journalism, sponsored by Poynter and The Associated Press
Newsrooms around the world — large and small, local and international — are experimenting with tools powered by artificial intelligence to connect their audiences to information, repackage their content in different languages and formats and augment their reporting and workflow capabilities.
But news audiences aren't always ready for it, says MJC director and Hubbard School researcher Benjamin Toff.
In the last year, news organizations including the San Francisco Chronicle and The Washington Post have created AI chatbots that allow readers to get information by communicating with virtual assistants that deliver AI-generated answers from previously published reporting, but a large plurality of the public says they have no interest in getting information from such tools.
Toff presented those findings to an audience of news and tech leaders, ethics specialists and researchers who gathered in New York last week for the Summit on AI, Ethics and Journalism, convened by The Associated Press and The Poynter Institute.
Poynter: Audiences are still skeptical about generative AI in the news
Toff, a leading expert in news avoidance, trust and audiences, joined the summit to share research from a study he conducted with The Poynter Institute about audiences' attitudes about AI in their news sources. The study is based on a nationally representative survey of 1,128 U.S. adults Poynter and the MJC fielded in March 2025. (More about the study.)
Among the study's other findings, which will be published in detail on the Journalism Center's website this month:
- Nearly one in five Americans say they are using generative AI tools like ChatGPT on a daily or weekly basis but 40 percent say they never have done so at all.
- More than a third of Americans feel ‘somewhat’ or ‘very’ fearful about AI compared to 23 percent who say they feel hopeful about it.
- A majority of respondents said disclosure is essential when news organizations use AI in their work.
- Only 20 percent of respondents said news organizations should avoid AI entirely, but most said they want newsrooms to establish ethics policies before diving into AI use.
The Poynter Institute's Alex Mahadevan, who directs the institute's MediaWise digital media literacy program, wrote about Toff's presentation and the study's findings within the larger context of the two-day summit.
The research builds on a series of focus groups Toff led last year in partnership with Poynter on these same subjects.
Read the full story: Audiences are still skeptical about generative AI in the news — Poynter
Explore more research from Ben Toff and the Minnesota Journalism Center team on our research overview.
The Minnesota Journalism Center supports a more vibrant, equitable, and sustainable ecosystem for journalism in Minnesota through educational initiatives, applied research and engagement with newsrooms and journalists across the state. Find out more about the center's work at mjc.umn.edu.