'An unstable situation': When AI practices medicine, who gets sued?

When AI practices medicine, who gets sued?

The shift toward digital-first care raises a question the legal system hasn't answered yet: when an artificial intelligence (AI) tool makes a clinical decision that harms a patient, who bears responsibility?

Medical Economics put that question to Richard E. Anderson, M.D., FACP, chairman and CEO of The Doctors Company and TDC Group, and he traced the problem back to something fundamental — the law has no framework for suing a machine.

Anderson used telehealth as a reference point. Traditional telehealth keeps a human physician in the loop. Push further in the digital direction and you arrive at what he calls AI telehealth — care delivered without a human clinician present.

That raises immediate questions about licensing, liability and accountability that remain unresolved.

"If you are a medical practitioner and you practice medicine without a license, that is criminal," Anderson said. "If you are a computer practicing medicine without a license, I assume it is illegal, but it seems to be happening."

When something goes wrong, Anderson expects the legal system to follow its established instincts. "Somewhere in that loop is a human physician, and that physician will be sued," he said. Suing AI developers — for example, Google, OpenAI, Anthropic or any number of smaller companies — is theoretically possible, and large plaintiffs' firms may pursue them for the size of their pockets. But those cases would take years to resolve, and any precedents established would apply only in a single venue and remain subject to appeal.

He pointed to electronic health records (EHRs) as an instructive parallel.

Despite being, as he put it, "a mixed blessing at best," EHRs have generated very little direct litigation. The suits go to the physicians and institutions that use them, and Anderson expects AI to follow the same pattern.