Key takeaways on Americans’ use and knowledge of and trust in AI 

While nearly all (98%) of U.S. adults have seen or heard information about artificial intelligence (AI) in the past 12 months, levels of use and knowledge of and trust in AI remain low, a report from Gallup and the nonprofit Special Competitive Studies Project finds. Based on an online survey of 3,128 U.S. adults conducted in April and May, 2025, Reward, Risk, and Regulation: American Attitudes Toward Artificial Intelligence found that 39% of respondents use AI sometimes, always, or often in their work, education, or personal lives. Here are a few of the report’s key findings: 

1. Only 8% of U.S. adults say they’re ‘very knowledgeable’ about AI 

Even among respondents who report using AI, only 13% said they were very knowledgeable about the technology. In terms of AI use, accounting for other factors, younger respondents and those with higher educational attainment were more likely to report using AI. 

2. Three in five U.S. adults do not trust AI to make fair, unbiased decisions 

Just 2% of respondents said they trust AI “fully” to make fair, unbiased decisions, while 29% said they trust it “somewhat” to do so. Sixty percent said they distrust it somewhat (40%) or fully (20%). People who use AI were twice as likely to say they trust AI as those who do not (46% vs. 23%), but nearly half of AI users (49%) still distrusted it.  

3. Majorities see AI as boosting productivity and U.S. economic growth 

Sixty-two percent of respondents expected AI to increase workplace productivity. In addition, 53% said it would boost economic growth; those with more wealth and educational attainment were more likely to say so. At the same time, 47% of respondents believed AI would destroy more businesses than it creates over the next 10 years, compared with 19% who believed it would create more businesses than it destroys. And 61% said AI would eliminate more jobs than it creates, compared with 12% who said it would create more jobs than it eliminates. 

4. Two in five support expanding AI training and education programs

When asked about AI-related policy options, respondents were most likely to express support for workforce training and education around AI development and use (41%), followed by increasing government funding for AI research (40%). There was significantly less support for providing tax incentives for AI companies (23%) or reducing regulation (17%). Half of respondents said “some” of those who lose their jobs because of AI will be able to learn new skills in a different industry, while just 3% said all workers displaced by AI would be able to do so.  

“Such mixed views suggest that economic outcomes alone do not drive public approval, and that perceptions of fairness, risk, and disruption could play a significant role in shaping support for future economic strategies regarding AI,” the report’s authors note. 

5. Younger respondents are less likely to say U.S. must be a global leader in AI

Most (79%) respondents agree that it is important for the U.S. to have more advanced AI technology than other countries. Yet, while 34% believe the U.S. is keeping pace with other countries, just 12% believe the U.S. is moving ahead, and 22% think it’s falling behind. Among respondents ages 18-29, only 26% see it as “very important” for the U.S. to have more advanced AI technology than other countries,” compared with 40% of those ages 30-39, 55% of those ages 40-64, and 64% of those 65 and older. 

The fact that AI users show greater trust in AI to make fair, unbiased decisions “suggests that greater adoption may build public confidence in AI technology over time,” the authors conclude. “Carefully managed deployment and exposure programs could help bridge the current confidence gap.”