In a survey of 2,150 adults, respondents aged 30 to 45 were far more open to AI girlfriends and boyfriends than those aged 18 to 29, and far more likely to have used one.
Generative AI, chatbots and digital relationships sound like quintessential Gen Z technologies. If any generation were going to embrace AI girlfriends and boyfriends first, you would expect it to be the youngest adults, the ones who grew up with a phone in hand and treat talking to software as ordinary.
A new survey finds close to the opposite.
Adults in their 30s and 40s were the most open
The survey, commissioned by AI Girlfriend Coach, a site that tests AI companion apps, was conducted through SurveyMonkey Audience on August 26 and 27, 2026. It collected responses from 2,150 U.S. adults, and 1,709 were kept for analysis after a quality screen removed a cluster of patterned answers. The sample was census-balanced for age and gender but not for region or income, so it describes the people surveyed rather than the American population as a whole, and it says nothing about Canadians in particular.
One question asked whether respondents would consider using an AI girlfriend or boyfriend in the future. The age gap was difficult to miss. 44.7% of adults aged 30 to 45 said yes, compared with just 12.5% of those aged 18 to 29.
The two groups in between and beyond followed the same curve. Respondents aged 46 to 61 were at 31.6%, and those aged 62 to 80 at 8.6%. The youngest adults, in other words, landed closer to retirees than to the age group just above them.
A note on labels: the survey’s report uses “Gen Z,” “Millennials” and “Gen X” as shorthand for its age brackets, not as generations determined from birth years. The groups that were actually measured are 18 to 29, 30 to 45, 46 to 61 and 62 to 80, and that is how they are described here.
Actual use follows the same pattern
The finding is not just a matter of hypothetical interest. Reported use lined up the same way.
Among respondents aged 30 to 45, 20.7% said they currently use an AI girlfriend or boyfriend app. Among those aged 18 to 29, 3.8% did. The 46 to 61 group sat between them at 12.5%, and the 62 to 80 group at 0.6%.
Counting current and past use together, 40.1% of the 30 to 45 group had used one of these apps at some point, compared with 16.8% of the 18 to 29 group.
The survey’s authors caution that reported use varied by device type even after screening, so the exact percentages should be read as describing this sample rather than as national adoption rates. What is consistent, on every measure the survey took, is the ordering: adults in their 30s and early 40s first, the youngest adults well behind.
Younger users were more likely to have stopped
There is a further detail in the youngest group that deserves a look, with the caveat that the numbers are small.
Only 31 respondents aged 18 to 29 had ever used an AI girlfriend or boyfriend app. Of those 31, just 7 were current users. The other 24 said they had used one in the past but no longer did. The report notes that this age group had the highest share of former users relative to current users of any group in the survey.
With a base of 31 people, that is an observation rather than a finding to build on. But it points in the same direction as everything else: younger respondents were not only less likely to have tried these apps, those who had tried them were disproportionately the ones who had walked away.
Why might adults in their 30s and 40s be more receptive?
The survey does not answer this, and it is worth being clear about why. It measured attitudes and reported use, not the reasons behind them. Any explanation is a possibility, not a result.
Several are plausible. People in their 30s and 40s are more likely to be established in long-term relationships, and the survey did find that engaged or married respondents reported higher use and openness than singles. They may have less free social time than younger adults, or different expectations of what companionship looks like. It is also possible that younger adults, who have more experience with AI tools generally, hold them to a higher standard or find the current products underwhelming. The survey cannot distinguish between any of these, and there may be explanations it did not measure at all.
The motivation was the same across ages
One thing the survey can say is that the appeal people described did not change much with age.
Asked what their main reason for using an AI companion would be, respondents in every age group most often chose having someone to talk to or feeling less lonely. Among those aged 30 to 45, 51.2% chose that answer, while 13.5% chose affection, romance or intimate interaction. The question was hypothetical and asked of everyone, so it reflects what people imagine the appeal would be rather than confirmed motives of users.
That matters because it separates two things that are easy to conflate. Adults in their 30s and 40s were the most open age group; separately, companionship was their most common stated reason. The survey does not show that one causes the other, and it should not be read as saying this generation is turning to AI because it is lonely.
Not the adopters you’d expect
Consumer technology usually spreads from the young outward. Social media did. Dating apps did. The assumption that AI companionship would follow the same path is reasonable, and this survey suggests it may be wrong, at least for now.
Among the people surveyed, the group most likely to consider an AI girlfriend or boyfriend, most likely to have tried one and most likely to still be using one was aged 30 to 45. The generation that grew up online was, on every one of those measures, among the least interested. Attitudes can change and the products are moving quickly, but the people most drawn to this particular corner of consumer AI are not the youngest people in the room.
The findings in this article come from AI Girlfriend Coach’s survey of 2,150 U.S. adults, conducted through SurveyMonkey Audience on August 26 and 27, 2026, and analyzed on 1,709 quality-screened responses. Respondents were U.S. adults; the results describe the surveyed sample.
