Americans wary of AI for medical care and insurance claims

How much artificial intelligence is too much when you visit your doctor? Although patients are generally open to healthcare professionals using AI to handle administrative and some medical tasks, they want humans wielding the stethoscope, a recent survey suggests.

A solid majority of Americans are comfortable with AI translating patient notes or scheduling appointments, either on its own or with some human oversight, according to the Pew Research Center, which polled more than 3,400 U.S. adults in June. More than half the respondents were also open to using AI to analyze medical scans and lab test results, provided a healthcare professional supervised the process.

But people draw the line when it comes to AI driving diagnoses or making insurance decisions, Pew found. Almost half of those surveyed said they don't want an AI model deciding whether to accept or reject an insurance claim, while 41% said they don't want AI making a diagnosis on its own.

Galen Stocking, associate director of science and society research at Pew Research Center, said people see value in applying AI to healthcare. Still, patients want to know when and how healthcare providers and administrative staff are using the technology.

"Overall, people say it's extremely important that their healthcare provider tell them if they are using AI," he told CBS News. 

"There is a lack of clarity on the side of the public," he added, noting Pew's finding that 46% of the survey respondents said they don't know if AI has ever been used in the course of them receiving care.

Indeed, 60% of respondents said they are uncomfortable with doctors using AI for tasks such as diagnosis and treatment; only 30% are OK with that, while 10% are unsure. 

"People are concerned about patient-doctor relationships and the security of their own health information," Stocking said. 

More broadly, Americans are split over whether AI will improve the quality of medical care. Some 37% of respondents said AI will improve health outcomes, while a nearly equal share think the technology could reduce the quality of care.