As AI beats doctors, regulators shouldn't force a human into the loop, JAMA piece says
| Source: THE DECODER
Tags: JAMA, medical AI, human-in-the-loop, FDA, Curai Health, Google AMIE, healthcare regulation
A JAMA opinion piece co-authored by a top bioethicist and a health AI CEO argues regulators should not mandate human oversight of medical AI, citing studies showing autonomous AI now matches or outperforms doctors on five core medical reasoning tasks — though nearly all evidence comes from simulations, not actual patient care.
Details
In a high-profile opinion piece in JAMA, Ezekiel Emanuel (UPenn bioethicist and former Obama healthcare architect) and Neal Khosla (CEO of Curai Health) argue that writing a doctor's final sign-off into AI regulation is a mistake. Their case rests on a run of studies since 2024 in which AI systems match or beat physicians on five core tasks: patient history, diagnosis, test selection, guideline-based treatment, and chronic disease management. The piece has notable conflicts of interest: Khosla runs an AI telemedicine company, and his father Vinod Khosla is an investor in both OpenAI and Curai Health. Physician groups like the AMA and American College of Physicians maintain that AI should assist — never replace — doctors. Medical professor Robert Wachter has publicly called AI-only care 'economy class' medicine. The authors counter that this ranking is unproven and argue regulators should wait for evidence before codifying human-in-the-loop requirements. The practical stakes are large: FDA and international regulators are actively drafting frameworks for autonomous medical AI. Google's conversational diagnostic system AMIE is cited as evidence of AI's clinical performance.