Suing the Ghost in the Medical Machine

Authors

• Jason R. Cheek, Samford University, Cumberland School of LawFollow

Abstract

Devices that autonomously diagnose medical conditions, with no human interaction, have arrived. These self-operating systems incorporate artificial intelligence (AI), and include machines that diagnose preventable blindness, as well as some that write radiology reports. These devices feature AI-integrated software to make diagnosis decisions traditionally made by humans who have received extensive medical training. Soon, AI-driven devices will autonomously make treatment decisions.

What happens when these autonomous medical systems harm a patient? At the moment, most would expect product liability law to provide a path to recovery. But AI oftentimes uses multiple complex inputs and approaches, including statistical probabilities, to reach its decisions. It is incredibly difficult, if not impossible, to determine whether AI-based products are defective using traditional product liability principles. It is a problem that should be addressed.

This paper proposes allowing developers of autonomous medical devices to be held liable under existing state medical malpractice laws. Currently, such statutes only apply to humans or healthcare institutions. By extending them to developers of autonomous devices, legislatures could achieve a reasonable middle ground that protects patients while encouraging development of these promising technologies.

Recommended Citation

Jason R. Cheek, Suing the Ghost in the Medical Machine, 29 SMU Sci. & Tech. L. Rev. 113 (2026)