Millions of people rely on AI-assisted tools every day — from digital personal assistants to smart home gadgets to devices in vehicles that assist with driving and navigation. Many of these AI-powered tools are increasingly being adopted by healthcare providers to assist with diagnostics, treatment recommendations, and administrative processes. However, a new study from the University of Minnesota School of Public Health (SPH) finds that most people do not trust their healthcare providers to use AI technology responsibly.

To conduct the study, researchers used a nationally representative survey conducted by the National Opinion Research Center between June and July of 2023, in which a total of 2,039 respondents were asked whether they trust their health system to use AI responsibly, and whether they thought their providers would protect them from potential harms from the use of AI tools.
The study, published in JAMA Network Open, found:
- A majority of people distrust their healthcare systems to use AI: 65.8% expressed low trust in their health care system’s ability to use AI responsibly, and 57.7% lacked confidence that their health system would make sure an AI tool would not harm them.
- Females were less likely than males to trust their healthcare system to use AI responsibly.
- People with experiences of past discrimination while seeking care were also less likely to trust healthcare systems with AI tools.
“This study underscores that healthcare systems are not currently seen as trustworthy AI users,” says Paige Nong, SPH assistant professor and lead author. “The rate at which new AI technologies are being adopted by healthcare systems is outpacing efforts to analyze and understand how patients perceive these tools. The stakes for patient trust are high and health systems should implement more transparent communication about their use of AI with an understanding that general patient trust, rather than specific aspects of the technology or health literacy, underlie patient perceptions.”
Funding support for the study was provided by the National Institute of Biomedical Imaging and Bioengineering.