Three Factors That Build Public Trust in Healthcare AI

Public acceptance of artificial intelligence in healthcare hinges on three distinct but interconnected factors, according to research published in JAMA Network Open that examined how patients form trust in medical AI systems.

The concept of "social license" — defined as dynamic, informal public acceptance that goes beyond regulatory approval — varies widely across technologies. While contactless payment readers operate with broad public trust, self-driving vehicles face ongoing skepticism. Healthcare AI falls somewhere in between, still earning its place in clinical settings.

Why it matters

As healthcare organizations invest billions in AI tools for diagnosis, treatment planning, and care coordination, understanding what drives patient acceptance becomes critical. Without social license, even technically sound AI systems may face resistance that limits their effectiveness and adoption, regardless of regulatory clearance.

The three pillars of acceptance

Researchers at the University of Queensland conducted workshops with 34 participants aged 31 to 60, all of whom had received care at Queensland healthcare facilities within the previous two years. The study, led by Dr. Clair Sullivan, identified three elements that shape social license for healthcare AI.

Relational engagement emerged as the first factor. Participants' trust in AI systems was directly influenced by their existing relationships with clinicians. Most preferred that their doctors inform them when AI tools are being used and how their data contributes to these systems. However, recognizing time pressures on physicians, some indicated they would accept this information from other healthcare staff, freeing clinicians to focus on direct patient care.

Structural support represented the second pillar. Participants expressed clear expectations for dedicated AI governance frameworks. Rather than viewing oversight as a constraint, they saw robust governance as essential for enabling AI's appropriate use in healthcare. These concerns intensified in scenarios involving long-term care, though patients in acute, high-stress situations prioritized immediate care over governance questions.

Performance reliability rounded out the framework. Participants demanded that AI-supported systems demonstrate consistent accuracy and deliver timely, effective healthcare guidance. While many expressed confidence in AI's technical capabilities, others raised concerns based on personal experiences where AI systems had produced errors.

Building sustainable integration

The researchers note their study is the first to systematically examine how healthcare consumers' perceptions either advance or impede social license for medical AI. They emphasize that social license rests on public trust and the expectation that AI will align with and promote public benefits.

The findings offer practical guidance for healthcare organizations working to integrate AI while maintaining patient-centered care principles. At a time when local, state, and national entities are refining AI governance frameworks, understanding these three factors becomes essential for sustainable implementation.

The research was first reported by the University of Queensland team and published in JAMA Network Open, where the full study is available at no cost.