July 10, 2026 – As artificial intelligence moves from the inbox to the concert hall, the publishing house and the studio lot, American attitudes toward the technology are a study in contrasts: most say they'd embrace AI working alongside a human artist, even as a clear majority draw the line at AI replacing that artist altogether.
New data from Angus Reid USA finds that when AI is described as working alongside humans — music, books, or television and film created “partly by AI and partly by humans” — a majority of Americans (50 to 54 per cent) say they'd be willing to read, listen to, or watch it. But when AI is the only creator, willingness falls by roughly ten points across every medium: just 39 per cent would be willing to watch a movie or TV show made entirely by AI, the lowest share of any category tested.
Look closer, and even that “willingness” comes with an asterisk. Among Americans open to fully AI-generated content, most qualify their answer, choosing “willing, but with some reservations” over “willing, with no real reservations” by roughly two-to-one. Genuine, unqualified enthusiasm for AI-only art is a minority position everywhere it was tested, ranging from just 11 per cent (TV/movies) to 13 per cent (music).
Willingness rises sharply once a human is part of the process
Respondents were asked how willing they'd be to consume six kinds of content: music, books, and TV/film, each tested in two forms — created entirely by AI and created partly by AI and partly by humans.
In every medium tested, adding a human hand to the process boosted willingness by 10 to 13 points. It's a signal that resistance to AI in the arts isn't really about the technology touching the work at all — it's about AI replacing the artist entirely. The share of Americans who are outright unwilling to engage falls sharply once a human is added back into the mix: 44 per cent say they wouldn't touch fully AI-made music at all, compared to just 36 per cent who say the same about music made with both AI and human input.
“Net willing” also hides real variation in how enthusiastically people say yes. The tables below break each figure into its full range of responses — and show that unqualified enthusiasm is rare, even among those who count as “willing.”
| When you think about AI and art or entertainment, would you be willing to read/listen/watch each of the following if it was created entirely by AI? | ||||
|---|---|---|---|---|
| Content | Willing, no reservations | Willing, some reservations | Not willing at all | Not sure |
| Music | 13% | 31% | 44% | 12% |
| Books/stories | 11% | 30% | 48% | 10% |
| TV/movies | 11% | 28% | 50% | 11% |
| When you think about AI and art or entertainment, would you be willing to read/listen/watch each of the following if it was created partially by AI and partially by humans? | ||||
|---|---|---|---|---|
| Content | Willing, no reservations | Willing, some reservations | Not willing at all | Not sure |
| Music | 17% | 37% | 36% | 10% |
| Books/stories | 15% | 36% | 39% | 10% |
| TV/movies | 15% | 37% | 39% | 10% |
The gap between “net willing” and genuine enthusiasm is the real story here: across every content type, roughly two or three times as many people are willing with reservations as are willing without them. Adding a human collaborator narrows that gap somewhat — unqualified enthusiasm rises from 11-13 per cent for AI-only content to 15-17 per cent for blended content — but even then, most “yes” answers come with a caveat.
Men consistently more open to AI-made art than women
Men are more open than women to AI-generated art in every single category tested, and the gap is both consistent and statistically significant — holding at 8 to 12 points whether the content is made entirely by AI or blended with human work.
Men are also more likely to give an unqualified “willing, no real reservations” answer, while women are more likely to say they're “not sure” — suggesting the gap isn't just about enthusiasm, but about a segment of women who simply haven't formed a firm opinion yet.
Older Americans are as open — or more open — than the youngest adults
Given how much more likely younger Americans are to use AI chatbots day-to-day, one might expect them to be the most receptive to AI-made art as well. The data tell a different story. In five of the six categories tested, adults 55 and older are as willing, or more willing, than adults 18-34 to engage with AI-generated content.
The middle-aged group (35-54) is, if anything, the most consistently skeptical, while the oldest cohort leads or ties for the top spot in five of six measures. For fully AI-made TV and film specifically, the youngest adults (35%) are actually the least willing of the three age groups, seven points behind those 55 and older (42%).
Chatbot familiarity beats age as a predictor of openness
If age doesn't explain who's most receptive to AI art, direct experience with AI does. Angus Reid USA also asked Americans how often, if at all, they have conversations with AI chatbots — tools like ChatGPT, Claude or Perplexity, used for discussing a topic of interest. Overall, 16 per cent of Americans talk to a chatbot at least once a day, 22 per cent do so at least weekly, and 44 per cent say they never do at all. Chatbot use itself skews young: 24 per cent of 18-34s talk to a chatbot daily, compared to just 7 per cent of those 55 and older, and nearly six-in-ten Americans 55+ (59%) say they never do.
Cross-referencing chatbot use with willingness to consume AI-made art reveals the sharpest divide in the entire survey — sharper than gender, and far sharper than age on its own. Willingness roughly doubles, and in some cases nearly triples, moving from Americans who never use AI chatbots to those who use them daily. Given that older Americans are still relatively unlikely to be regular chatbot users, their comparatively high openness to AI-made art suggests something other than day-to-day familiarity with the technology is driving their comfort level — while for younger and middle-aged Americans, comfort with AI in one context (conversation) appears closely tied to comfort with it in another (creative work).
Methodology and downloads
Angus Reid USA conducted an online survey from June 22 – 26, 2026, among a representative randomized sample of 2,007 American adults who are members of Angus Reid Forum USA. For comparison purposes only, a probability sample of this size would carry a margin of error of +/– 2.0 percentage points, 19 times out of 20. Discrepancies in or between totals are due to rounding. The survey was self-commissioned and paid for by Angus Reid USA.