A major new poll suggests that American creators and makers are more worried about the erosion of their autonomy than the tools themselves. While the industry often debates whether artificial intelligence will go rogue, the public is far more concerned about losing their livelihoods and their ability to think for themselves. The data indicates that while people acknowledge AI’s utility, they remain deeply resistant to its integration into their personal workflows.
The hope for healing, the fear for the mind
A comprehensive survey conducted by Anthropic involving nearly 52,000 Americans reveals a clear hierarchy of expectations and anxieties. The respondents’ primary hope for the technology is medical breakthroughs, specifically curing serious illnesses like cancer and Alzheimer’s disease, with 48 percent of participants selecting this option. Helping individuals with disabilities ranked second at 36 percent, while general technological progress and simplifying daily life were tied at 23 percent each. Conversely, the idea of AI serving as a therapist or a cure for loneliness, a concept frequently pitched as a future opportunity, was the least desired outcome.
The anxieties are dominated by economic and cognitive threats. Sixty-four percent of the population fears that AI will cost them their jobs. This concern is universal, ranking as the top worry in every single state surveyed and intensifying among those with higher levels of education. Closely behind is the fear of cognitive dependency—the loss of independent thinking—which plagues 56 percent of respondents. Misinformation concerns round out the top three at 52 percent. Notably, only 15 percent of Americans trust AI companies to make the correct decisions regarding how the technology is built and deployed.
Experience mitigates the dread
Anthropic conducted this study as part of its “Anthropic Public Record” series, published between November and December 2025. The poll, carried out by YouGov, covered 51,993 Americans aged 16 and older across all 50 states, the District of Columbia, and Puerto Rico. It marks the first time the company has released data gathered from the general public rather than just its own users of the Claude chatbot.
Despite the apprehension, there is a pragmatic recognition of capability. Approximately 75 percent of respondents rated AI as at least as good as humans at research tasks. Yet, a significant disconnect remains between perceived ability and workplace acceptance. A majority of Americans do not want AI involved in their jobs, even on tasks where the technology excels.
On most tasks, a majority of Americans did not want AI involved in their jobs, and even on the tasks they rated AI most capable—such as research and data analysis—nearly half of respondents said they want no AI involvement in their own work.
The survey highlights that acceptance correlates with familiarity. Those who use AI daily at work are significantly less worried about job loss (54 percent) compared to non-users (70 percent). Practical experience appears to temper the fear, likely because it demonstrates both the utility and the limitations of the technology. This finding was reinforced by a separate qualitative study of 81,000 Claude users, where job loss and cognitive dependency also topped the list of concerns.
Anthropic notes that these fears echo historical anxieties surrounding automation, smartphones, and social media. In general, the public worries more about humans misusing the technology than the scenario of AI turning against humanity. Regarding cognitive dependency, the fear is often theoretical; of those worried about losing their ability to think independently, only about one in five would genuinely struggle if the technology vanished tomorrow. Ironically, among those not worried, roughly a third would feel a genuine disruption.
Workers in the arts, design, and education sectors are the most concerned about cognitive atrophy. Educators, in particular, report observing this decline in their students 2.5 to 3 times more often than the average respondent, according to earlier studies.
Key takeaways
- Medical cures are the top hope for AI, while job displacement and loss of independent thought are the dominant fears.
- Practical, daily usage of AI tools significantly reduces anxiety about employment security compared to non-users.
- Despite acknowledging AI’s competence in research and analysis, most Americans still reject its presence in their own workplaces.
- The public fears human misuse of AI far more than the technology itself going rogue or becoming sentient.
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.
