Georgia Tech pair press patient-first medical AI liability

Georgia Tech researchers Gennie Mansi and Mark Riedl analyzed 31 US legal cases and reported harms from medical AI tools and argue, in Nature Health, that the current accountability frame points at the wrong person. Physicians sit at the centre. The harms, in their reading, mostly do not originate there.

Patient care, they write, "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." Named AI systems in the case pool include nH Predict, the Oregon DHS Safety at Screening Tool and the Babylon Health Triage System.

"In response, patients have had no option but to seek legal recourse for harms," the authors write.

Mansi and Riedl argue for two moves. Rework liability rules so they follow the chain of decisions the tools sit inside: insurers, state health departments, care facilities. And build software for "advocates, such as patients' lawyers, who provide critical legal expertise and practically support recourse for patients."