Overall, the report is based on over 10,000 responses nationwide from April 2025 to August 2025. (11/11)

More results, including trends and additional information about AI adoption and usage are in the report: civiqs.com/reports/2025...

As we write: "Establishing trust in the helpfulness and safety of AI will require that people have successful interactions with AI technology, as companies work to address the concerns that people express about AI today." (10/)

The concerns that Americans express about the potential risks of AI suggest concrete steps to reduce AI skepticism and increase public confidence: for example, improving the reliability of AI output, stronger safeguards, and more robust oversight would all help. (9/)

The breadth of issues raised reveals that regardless of their level of AI use, people are very aware of its impacts. The report explores each set of concerns and includes verbatim quotes from the poll. As one respondent put it: "Terminator did not end well for the humans." (8/)

We wanted to go deeper on why people were worried about AI, so we asked respondents to answer in their own words: “In what ways do you think Artificial Intelligence is unsafe?” The responses clustered into 11 core ideas, including hallucination, misuse, and lack of regulation. (7/)

People who use AI in their personal lives even just a few times a month are also positive about the technology. But, people who do not use AI at all -- which are 51% of adults -- are strongly negative about AI, with 55% saying AI is a bad thing, and 22% saying AI is good. (6/)

An encouraging finding is that the most frequent AI users are overwhelmingly positive about the technology. Among people who use AI daily in their personal lives, 72% say that AI is a good technology, and only 12% say it is bad. However, this is only 6% of users. (5/)

The results also uncover an alignment issue: what AI is doing is not what many people say they want. Overall, more say that AI is a bad thing (44%) than a good thing (35%); and more believe AI is having a negative (49%) rather than positive (18%) impact on the world. (4/)

The widely-reported public trust problem in AI is real. Americans are more negative than positive about AI, in ways that go beyond any single company or use case.

Only 33% of Americans say they generally trust AI to be helpful, and just 19% trust AI technology to be safe. (3/)

Usage of AI is increasing, measured both by first-time use and frequency of interaction. Adoption is being led by younger Americans and those with higher levels of formal education. Yet sentiment towards AI has worsened during 2025. (2/)

The early results have been fascinating: Over time, you can see which issues come and go and which have staying power. It gives you a view directly into the media bubbles of people who are very different from yourself.

Here again is the link to get Immediate Reaction in your inbox! bit.ly/4lP1yKP

The daily polling is already part of what Civiqs does. What's new is that we've fully integrated AI into summarizing the responses into a clear and succinct readout.

We pull out the top stories that people mention, and describe how Democrats, Republicans, and Independents each react to them.

With how quickly current events are developing—and how fragmented the media landscape has become—we survey 400-500 Americans nationwide every day to answer in their own words the question "What's your immediate reaction to something you've seen or heard in the news lately?"

Here you go

Another significant trend is among Hispanic/Latino voters, who have gone from net +9 favorable views of the Democratic Party in 2015 to net unfavorable by 20 points in 2025. (3/3)

For over a decade, Democrats have been steadily losing support from young voters, especially Independents age 18-34, under both Trump and Biden. (2/3)