Trust Splits by Task: Admin vs Diagnosis

A national public tends to accept AI for administrative and back‑office healthcare tasks (scheduling, billing, record‑search) but is much more reluctant to accept AI for core clinical decisions like diagnosis, prognosis, and treatment selection. That task‑based trust gradient will shape incremental adoption paths and what regulators and hospitals prioritize for pilot programs. — If publics accept AI for non‑clinical tasks but reject it for diagnostic decisions, policymakers and health systems will face pressure to confine early deployments to paperwork and automation while putting stronger rules around clinical autonomy.