As AI beats doctors, regulators shouldn't force a human into the loop, JAMA piece says

Authors writing in JAMA suggest that future autonomous AI may eventually surpass human-led diagnostic teams in clinical reasoning. The piece argues that regulators should avoid mandates requiring a doctor's final approval for every AI decision, claiming such oversight could hinder performance. However, the authors note that these projections are based primarily on simulations rather than actual patient care. The perspective challenges current regulatory frameworks that prioritize human-in-the-loop requirements as artificial intelligence systems grow increasingly sophisticated in medical settings.