More than 60% of Americans have at least some confidence that news organizations will provide accurate information about the midterm election results, but only 32% have high levels of confidence, according to a poll for the Medill Local News Initiative and the Alliance for Trust in Media.
Additionally, 58% expressed at least some confidence that news outlets will explain what early election results mean, while the same number had at least some confidence that these entities would wait until enough votes are counted and the outcome is conclusive before declaring winners.
Just over half of respondents have at least some confidence that news organizations and journalists will thoroughly investigate claims of voter fraud.
With Election Day approaching next Tuesday, Americans across the country will be voting in U.S. Senate, U.S. House, gubernatorial and a number of other state and local elections.
General trust in media and the political system have both declined significantly in recent decades, and President Donald Trump has repeatedly questioned election integrity. These survey results aimed to measure Americans’ perspectives heading into an election that’s seen more than $10 billion in ad spending.
This poll is part of a new quarterly series in a partnership between the Alliance for Trust in Media and the Local News Initiative at the Medill School of Journalism, Media, Integrated Marketing Communications at Northwestern University. The polls are being conducted for Medill and the Alliance by NORC at the University of Chicago. For more information about the partnership and the first part of the results focused on Americans’ general levels of trust in news organizations and journalists, click here.
Among the poll’s key findings: A plurality of Americans believe that news organizations and journalists manipulate the news for political and commercial purposes.
“It certainly creates a lot of challenges around some of the functions that I think we tend to ascribe to journalism — helping people make sense of what’s true and what’s factual, what’s not; in holding leaders to account when they are disseminating false information themselves,” said Benjamin Toff, the Alliance’s head of research and the director of the Minnesota Journalism Center at the University of Minnesota, where he serves as an associate professor. “The degree to which people don’t believe journalists have the public’s interest at the center of what they do, that definitely creates some problems in terms of how people navigate the information environment.”
There were partisan differences among respondents, with self-identified Democrats more confident than self-identified Republicans that news organizations and journalists will provide accurate election coverage. Democrats and Republicans include respondents who lean toward each party.
Even still, at least a plurality of Republicans say they had at least some confidence that news organizations would provide accurate information about election results, explain what early results mean and wait until a conclusive outcome before declaring a winner. Investigating voter fraud claims was the exception, with 57% of Republicans saying they had not too much or no confidence that news outlets would do so thoroughly.
This suggests there’s a path to rebuild higher trust, Toff noted.
“Certainly, some people I think are hostile that news organizations and journalists will never do any of these things,” he said. “But I think partly what you’re seeing here is even among Republicans, there’s still some willingness to give news organizations and journalists some benefit of the doubt.”
The survey also asked Americans how much they trust news organizations, social media platforms, political campaigns and themselves to detect AI-generated or manipulated content about the election.
Notably, just 20% said they had “quite a bit” or “a great deal” of trust in their own ability to detect AI content, followed closely by 19% who said the same about news organizations. Trust was even lower for campaigns (10%) and social media platforms (8%).
“Our findings signal that for a lot of people, there isn’t a gatekeeper they trust to spot AI fakes, so they are relying on themselves,” said Stephanie Edgerly, Medill’s associate dean of research and professor who helped lead the poll. “That’s very understandable, but I want to underscore that confidence isn’t the same as skill. And it’s meaningful that trust in news organizations is essentially tied with people’s trust in themselves.”
That’s one of the positive takeaways Edgerly said she sees from the data — that people are differentiating news organizations from political campaigns and social media platforms.
“Maybe that should be obvious, and not a huge win for news organizations,” Edgerly said. “But I’d ask newsroom leaders: How can you position yourselves to help the public keep an eye on social media sites and campaigns? People don’t trust social media sites to handle AI content, even though they spend time there. And they don’t trust the campaigns, even though campaigns are targeting them with ads and mailers.”
“So how can you help people make sense of that environment? That’s where news organizations have a potential leg up: being on the side of the public, and helping them make sense of election content in a way that social media companies and campaigns can’t.”
About the survey
The Alliance for Trust/Medill Trust Tracker was conducted by NORC at the University of Chicago from October 8 to 12, 2026, among a nationally representative sample of 1,057 U.S. adults ages 18 and older. Respondents were drawn from NORC's AmeriSpeak® Omnibus® Panel and completed the survey online or by telephone, in English.
The survey was funded by the Local News Initiative at Northwestern University's Medill School of Journalism, Media, Integrated Marketing Communications and the Alliance for Trust in Media. Questions were developed jointly by Medill and Alliance researchers. The polling project partnership was formed by Tim Franklin, professor and John M. Mutz Chair in Local News at Medill, and Eric Schurenberg, founder of the Alliance.
AmeriSpeak is a probability-based panel, meaning households are randomly selected to join rather than volunteering. NORC samples U.S. households from its National Sample Frame and recruits them by mail, telephone and in-person visits. The panel covers about 97% of U.S. households. Households without internet access can take part by phone.
To ensure the results reflect the U.S. adult population, the data are weighted to U.S. Census Bureau benchmarks for sex, age, education, race and ethnicity, and region.
The margin of sampling error for the full sample is ±4.1 percentage points at the 95% confidence level, accounting for the design effect from weighting. Margins of error are larger for subgroups, such as age groups or political parties. Reported differences between groups are statistically significant at the 95% confidence level. In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias. The cumulative response rate (AAPOR RR3) is 4.5%. For more about AmeriSpeak, visit AmeriSpeak.norc.org.
"High confidence" or “high trust” combines "a great deal" and "quite a bit"; "little or no confidence" or “little or no trust” combine "not too much" and "none." Party identification was measured on a seven-point scale. Democrats (strong, not so strong and leaning Democrats) and Republicans (strong, not so strong and leaning Republicans) include those who lean toward each party; independents are those who do not lean toward either party.
