The Pulse

Pew Brief Revisits Americans’ Unease With Healthcare AI

A Pew Research Center brief presents findings from a survey in which 60% of U.S. adults said they would be uncomfortable with their doctor using AI for diagnoses and treatments. The seven-slide presentation draws on research published in October, rather than a newly conducted survey.

AI.info Team · · fact-checked · against 8 sources ↓ · Published 12 Nov 2026

Six in 10 U.S. adults would be uncomfortable if their doctor used artificial intelligence for tasks such as diagnosis and treatment, according to a Pew Research Center survey. The center published a seven-slide brief on November 11 presenting findings from its study of attitudes toward AI in healthcare.

The presentation draws on a report published in October, not a newly conducted survey. That report was based on responses from 3,488 U.S. adults surveyed June 22-28, 2026. All participants belonged to Pew’s American Trends Panel, and the center says the survey represents the full U.S. adult population.

Views depend on the task

Respondents were more receptive to some uses of AI when a person would review the result. For patient translation, 52% said they were open to AI doing the work if a human conducted a final review, while 26% were open to AI doing it on its own; 12% were not open to AI doing it.

For medical scans, 60% were open to AI analyzing them with a person’s final review, compared with 6% who were open to AI working alone and 26% who were not open to its involvement. Insurance claims drew less support: 46% said they were not open to AI making a decision on a claim, while 38% were open to it with human review and 5% without it.

Outcomes and privacy

When asked whether AI used for things like diagnoses and treatments would affect patient health outcomes, 37% said outcomes would improve, 36% said they would worsen and 26% expected little difference. Those answers concern respondents’ expectations, rather than measured changes in patient health.

On information security, 51% said AI would make the security of patients’ personal health information worse, while 8% said it would make security better. The survey also found that 47% expected AI to worsen the patient-doctor relationship, compared with 10% who expected an improvement.

Patients want to be informed

A separate analysis of the same survey, published in August, found that 72% of U.S. adults considered it extremely or very important for a doctor or other healthcare provider to tell them if AI was being used in their healthcare. The question covered disclosure in their own care, rather than whether they supported AI use in every task.

The same analysis found that 81% wanted a provider or office to tell them if AI was used to analyze medical scans. It also found that 63% wanted more say over whether AI was used in their healthcare, while 21% were comfortable with how much say they had and 16% were unsure.