A Morgan Stanley survey of roughly 2,000 Americans finds that favorable views of artificial intelligence are substantial but not uniform across the population. Overall, 44% of respondents expressed positive views of AI, 28% said they felt negatively about the technology and 25% were neutral, producing a net favorable reading of 16%.
Frequency of use correlates strongly with sentiment. Weekly users were the most optimistic group: 74% held favorable views while just 7% were negative, yielding a net score of 66%. By contrast, Americans aged 55 and older were the only demographic group recorded in negative territory, at 13%.
The survey showed a clear age gradient in net favorability. Net positive sentiment peaked among 25-to-34-year-olds at 46% and remained high among 35-to-44-year-olds at 38%. It then declined to 19% among 45-to-54-year-olds and to 13% among 16-to-24-year-olds.
Income and gender also aligned with differing attitudes toward AI. Households earning $100,000 or more reported a 26% net favorable reading, while those earning under $50,000 registered a 7% net favorable score. Men showed a 28% net favorable score against a 6% reading for women.
Political identification tracked with sentiment as well: conservatives posted a 31% net favorable score, those in the middle 17%, and liberals 4%.
Among respondents who viewed AI positively, the leading reasons were practical and user-focused. Saving time and improving access to information were each cited by 53% of those with favorable views. Productivity gains were mentioned by 37% and a positive personal experience by 35%. Broader societal benefits such as improving healthcare or supporting economic growth trailed at 24-25%.
Negative views were driven more by concerns than by direct experience. Among consumers expressing unfavorable opinions, 51% pointed to fraud and harmful use, 45% cited misinformation, 44% mentioned insufficient safeguards and 42% raised data privacy concerns. Worries about nationwide job losses were present for 34% of negative respondents, while only 18% worried about the impact on their own employment. Opposition to data centers being built nearby was cited by 25%, and just 6% attributed negative sentiment to a personal negative experience with AI tools.
Usage patterns reported in the survey offer additional texture. In April, 22% of consumers said they used AI daily for personal purposes, and roughly half reported using AI at least weekly - a level described as stable since January. Research and learning topped the list of personal use cases at 42%, followed by recipe searches at 27%.
At work, adoption was lower overall but concentrated among white-collar roles. Twenty-nine percent of employed respondents reported using AI for work tasks in the past month. White-collar workers led adoption at 39% versus 22% for blue-collar workers, and white-collar respondents reported higher weekly usage at 62% compared with 43% for blue-collar workers.
Among those using AI on the job, information gathering was the most cited activity at 68%, up from 61% the prior month. Writing and editing tasks were cited by 61%, down from 66% the prior month.
The survey encompassed approximately 2,000 Americans aged 16 to 75 and carries a margin of error of 1.8% at the 90% confidence level.
Key takeaways:
- Overall net favorability toward AI stands at 16%, but sentiment varies widely by demographic and usage.
- Frequent users are highly positive while older Americans show net negative views.
- Practical benefits drive positive sentiment; concerns about fraud, misinformation and privacy drive opposition.