Most Americans use AI but still don’t trust it

Key findings (YouGov AI trust survey, Dec 2025)

• 35% of U.S. adults use AI tools at least weekly; Gen Z leads at 51%.
  
• Only 5% of Americans say they “trust AI a lot”; 41% express distrust.
  
• No industry earns a net-positive trust score — trust is lowest in finance (19%) and healthcare (23%).
  
• 68% of respondents wouldn’t let AI act without specific approval.
  
• 77% are concerned that AI could pose a threat to humanity.
  
• Trust gaps are generational — younger adults show higher adoption and more optimism, but skepticism persists across all age groups.
  

Artificial intelligence is now part of everyday life for many Americans, but most still don’t trust it. New data from a December 2025 YouGov survey shows skepticism is especially high in finance, healthcare, and other sensitive industries. While younger Americans are more open to the new technology, all generations show concerns over how much they can trust AI and whether it will eventually pose a risk to humanity.

How many Americans of each generation use AI weekly?

AI tools are becoming embedded in daily life. According to the YouGov AI survey, 35% of US adults report using AI at least once a week in the past year, while another 28% say they’ve used it less often. Just 30% of Americans say they haven’t used AI at all.

The data also reveals a significant generational divide, with 51% of gen Z reporting weekly usage compared to only 29% of gen X and 25% of baby boomers. This suggests AI tools have already become a regular part of life for many younger Americans, though uptake among older groups remains limited.

Trust in AI is low and declining

Despite widespread use, Americans remain cautious about AI. When it comes to making recommendations or providing information, just 5% say they trust it “a lot”, and 26% trust it “somewhat”. Meanwhile, 41% express distrust (19% “not much,” 22% “not at all”), and 23% are neutral. This suggests that while many are using AI chatbots to get quick answers to their queries, most people are not convinced they’re getting accurate answers.

Trust is even lower when it comes to AI system taking action. In 2025, many in the AI industry raised expectations that autonomous AI agents could start independently accomplishing tasks. Americans are skeptical. Fewer than one in five (18%) would trust an AI system to make a decision or take an action, even “somewhat.” This compares to a 53% who do not trust these systems. Even within gen Z the mistrustful outweigh the trustful 43% to 26%.

Although trust in AI is low, it appears to be getting even worse. When asked how their trust in AI has changed over the past year, only one in five (21%) say it has increased. More (25%) say it has decreased while about half of all respondents (47%) say they trust AI about the same as they did a year ago.

This indicates that there has been no overall gain in public trust, and in fact, slightly more Americans report a decline in trust than an increase. Generational patterns again emerge. Gen Z is the most likely to report increased trust (33%), compared to 23% of millennials, 19% of gen X, and just 14% of baby boomers.

Which industries do Americans trust AI in? None

While people may use AI in their daily lives, their willingness to trust it varies greatly depending on the industry. No sector achieves a net positive trust score, meaning there is no sector in which more Americans trust AI than distrust it.

The least distrusted sector is leisure and entertainment, which is balanced with 31% of Americans on either side, for a net trust of 0. Dining (-4 net trust) and tech (-4) are comparatively more trusted with AI than other business sectors.

However, in sensitive or high-stakes sectors like finance, healthcare, and automotive, trust declines sharply. In financial services, only 19% of Americans trust AI, while nearly half (48%) say they don’t, for a net trust of -29. In health and personal care, 23% express trust compared to 46% who express distrust (-23).

What tasks won’t Americans trust AI with?

While people may use AI tools for various tasks, most Americans draw hard lines on where it can be used. Asked what they would not trust AI to do—even in principle—respondents identified a range of roles.

Significant majorities say they would never trust AI to act on their behalf without reviewing each action (68%) or to make political decisions (63%).

Even in areas where AI is commonly applied—like travel planning or sports officiating—more than a third of Americans say they would not trust AI.

Just 10% of respondents say there’s no task they wouldn’t trust AI with, and among the 2% who wrote in custom responses, most expressed complete distrust in AI for any task.

Most Americans are concerned AI could threaten humanity

Beneath sector-specific concerns lies a deeper unease about AI’s long-term implications. More than three quarters (77%) of Americans are concerned that AI could pose a threat to humanity, including 39% who are very concerned. Only 17% say they are not very or not at all concerned.

AI use in the US may be increasing, but greater exposure has not led to greater trust, especially in sectors that involve personal data, finance, or health. For businesses and policymakers, this means adoption alone is not enough: they must clearly communicate why AI should be trusted.

Methodology: YouGov polled 1,287 US adults online on December 4, 2025. The survey was carried out through YouGov Surveys: Serviced. Data is weighted by age, race, gender, education, and region to be representative of all US adults (18 years or older), and reflects the latest population estimates from the US Census Bureau. For the question “"In general, how much, if at all, do you trust AI (Artificial Intelligence) systems to go beyond providing information and make decisions and/or take actions?”, a separate survey was conducted on December 8 of 1187 Americans.