Loading...
AI is increasingly used in clinical settings, and its success depends in part on whether patients trust it and are willing to accept it as part of their care. However, little is known about what drives patient trust of AI during clinical encounters.
To explore this, researchers asked 3000 U.S. adults to imagine a clinical scenario in which they had a rash diagnosed by an AI system at a medical facility. They were then shown two different versions of the encounter and asked which inspired greater trust. Each version varied by whether a clinician was present, how the AI system was known to perform relative to clinicians, whether the system was vetted by well-regarded organizations, and whether it was trained on a dataset that represente…