Nature Health: 31 US AI-care cases move liability past doctors

A new Nature Health analysis of 31 US legal cases involving AI tools in healthcare argues that liability frameworks centered on the treating physician miss where much of the harm is actually landing. Georgia Tech's Gennie Mansi and Mark Riedl write that patient care "relies on a complex web of stakeholders—physicians, state health departments, health insurers and care facilities, among others—and many AI tools deployed across their healthcare delivery negatively impact their care."

The reframe the authors press is from "physician-centred to patient-centred accountability approaches." Patients, they write, "have had no option but to seek legal recourse for harms." Mansi and Riedl offer two paths: changing liability structures so they reflect the full stakeholder chain, and designing tools that let patient advocates, especially lawyers, actually mount that recourse.

The abstract does not name which 31 cases were analyzed, identify particular defendants, or report win-loss data. It positions the argument as an analysis of legal cases and reported harms, aimed at lawyers and technologists rather than at physicians alone.