AI is gaining usage. But there are some things Americans would be comfortable having AI do for them — and plenty of others they aren’t, recent CBS News polling found.
Given a list, the sorts of things Americans are comfortable with AI doing tend to be impersonal or even boring tasks, as well as those that less directly impact them, like proofreading or searching online.
But they’re less comfortable with AI doing things that have more immediate and perhaps more consequential impact on them: Making medical diagnoses, doing their taxes, handling their finances or driving their taxis.
It is newer technology, but there are only slight differences by age on these views.
If we look closer at the differences between expressed comfort and discomfort, for some of those items (for which a majority are uncomfortable) the difference is dramatic.
In any case, big majorities believe AI is going to reduce the number of jobs available in the U.S.
More generally, Americans don’t have much confidence that AI companies will ensure AI will be used in appropriate ways.
Americans, themselves, report using it more compared to last year. A majority of Americans now say they use AI for something — mostly for their own use and not at work. The reported increase in use spans age, education and racial groups.
Thinking about the government’s role, more continue to feel that government policy ought to be to restrict the use of AI, not to promote it. (This perhaps naturally follows from the views on job impact. People who think it’ll decrease jobs are relatively more likely to say restrict it.)
And there’s some collective skepticism about whether the military ought to use it for analyzing military and intelligence data.
This might be in keeping with the general approaches they’d apply for themselves, too. Those who wouldn’t want A.I. handling their finances or driving their taxi are also more likely to say the military shouldn’t use it for analysis, either.
This CBS News/YouGov survey was conducted with a nationally representative sample of 2,500 U.S. adults interviewed between March 16-19, 2026. The sample was weighted to be representative of adults nationwide according to gender, age, race, and education, based on the U.S. Census American Community Survey and Current Population Survey, as well as 2024 presidential vote. The margin of error is ±2.2 points.