Defining Responsibility as AI Takes on Greater Roles in Patient Care

Artificial intelligence is rapidly entering hospitals, but the shift toward more autonomous tools raises unanswered questions about legal responsibility. Jianing Qiu argues that current liability structures—covering clinicians, institutions, manufacturers and regulators—do not fit AI systems that blend human and machine judgment. To address this, she outlines a seven-level taxonomy based on autonomy, automation and operational scope, mirroring grading systems for aircraft and self-driving cars.

Levels range from purely informational tools that merely display data to fully autonomous systems that manage entire treatment episodes. By assigning clear risk categories, regulators can set evidence standards, hospitals can design consent protocols, and courts can better assess negligence or defect claims. The approach seeks to close liability gaps and encourage safe adoption of advanced medical AI.