Two-thirds of patients distrust AI in health care

A majority of U.S. adults have low trust in their health care systems to use artificial intelligence (AI) responsibly, according to a new study published in JAMA Network Open. The study underscores a need for improved communication about AI-driven tools in patient care.

Led by researchers from the University of Michigan and the University of Minnesota, the study found that 65.8% of U.S. adults expressed low levels of trust in their health care system’s ability to use AI responsibly. Additionally, 57.7% had low trust that their health system would ensure AI tools do not cause harm.

“The adoption of AI technology in health care is outpacing efforts to analyze patient perspectives,” wrote the study’s authors, Paige Nong, PhD, and Jodyn Platt, PhD. They concluded that “low trust in health care systems to use AI indicates a need for improved communication and investments in organizational trustworthiness.”

The survey, conducted in mid-2023, analyzed responses from a nationally representative sample of 2,039 U.S. adults. Results showed significant variations in trust levels based on demographics and previous health care experiences.

Key takeaways

• Patients with higher trust in their health systems were more likely to believe their providers would protect them from AI-related harm.
• Women were less likely than men to trust their health care systems to use AI responsibly.
• Past experiences of discrimination in health care were associated with lower trust in AI-related decision-making.
• Neither AI knowledge nor health literacy significantly impacted trust levels. This suggests that skepticism could stem from broader concerns about health system reliability, rather than from a lack of understanding.

The study’s findings emphasize the importance of proactive patient engagement and communication regarding AI implementation in clinical settings. Patients who already trust their physicians may be more open to AI-assisted decision-making, but skepticism remains high among the general public.

As AI integration into clinical practices becomes more prevalent, physicians may need to address patient concerns directly, ensuring they understand how the AI tools are being used, and what safeguards are in place to protect them.

The researchers recommend that future studies track patient trust over time, assessing whether familiarity with AI improves perceptions. With AI-driven health care innovations on the rise, the study serves as a critical reminder that technology alone cannot improve patient outcomes — trust and communication must be part of the equation.