AI's M.D. Bid

AI tools are already speeding drug development, uncovering new uses for existing medicines and helping patients manage chronic conditions. But as this frontier expands, how far is too far? According to new research from Sacred Heart University, Americans are open to AI in health care, but only with strict conditions related to transparency, control and trust.

“Patients are generally open to the use of AI in aiding, not replacing, doctors,” says Foluke Omosun, assistant professor and director of Sacred Heart’s strategic communication & public relations program. “All they’re asking for is transparency. In other words, they use it themselves for health information seeking, but they want disclosure and privacy protection when it’s used as part of their health care management.”

Omosun recently collaborated with Anna Price, SHU professor of public health, to conduct a national survey of 1,500 U.S. adults examining how Americans use, perceive, and trust AI in health care.

The data show a clear boundary around the exam room. When AI plays a role in their care, respondents were far more comfortable with low-stakes applications, such as scheduling (39.6%) and reminders (41.0%), compared with diagnosis (24.7%) or surgery (12.7%). “Patients are understandably cautious about how their personal health information is used and protected,” Price says.

In other words, although respondents accept that AI is part of modern health care, they hesitate when the technology enters the exam room.

“It is also critical to recognize that the usefulness of AI depends on users’ ability to interpret and evaluate the information it provides,” Price says. “AI should be viewed as a supplementary tool that can support, but not substitute for, clinical expertise and judgment.”

Recent cases show AI testing the boundaries. Earlier this year, Apple removed an app called Eureka Health: AI Doctor, which marketed itself as an authoritative source of medical advice at scale and urged users to “become your own doctor.”

“These tools are not currently a substitute for professional medical advice,” Omosun warns. “They have been found to have biases and assumptions, so any information needs to be double-checked for accuracy and relevance to an individual’s specific health situation. AI also tends to give overwhelmingly positive responses, which means users need to question and critically evaluate the output rather than accepting it at face value.”

Surveys from Gallup consistently rank nurses, pharmacists, and doctors among the most trusted professions in American life despite broader skepticism toward institutions. That means patients are far more willing to accept AI when it is overseen and explained by a doctor.

Indeed, survey respondents say they want transparency when AI is being used in their care—and want the option to opt out. When asked who should regulate these tools, many pointed not to government agencies or professional medical associations but to local clinicians themselves, a preference that reveals “trust in providers,” Omosun says.

Price has noticed emerging discussions following the survey, including the technology’s environmental footprint—“particularly the energy and water demands associated with large data centers.”

In the meantime, as late-night AI consultations become a common first stop for patients, the doctor’s office—at least for now—retains the final word.

Want to hear more from SHU? Subscribe to our Pioneer Times newsletter to get the latest updates delivered right to your inbox.