ARPA-H's $62.7M Autonomous AI Bet Raises a Liability Question

HHS is betting on autonomous AI. On Sept. 9, the department's Advanced Research Projects Agency for Health selected six organizations to build an AI agent that functions as a round-the-clock member of a heart failure patient's care team, managing care between visits and escalating to a clinician only when needed, according to Becker's Hospital Review. The work runs under a four-year, $62.7 million contract.

The project pushes past AI tools that only suggest actions to a clinician. An autonomous agent would act on its own, then loop in a human when a case demands it. That shift is what worries health system leaders. The question they keep returning to is accountability: when an agent makes or delays a clinical decision, who carries the liability, the vendor, the health system, or the clinician overseeing it.

No clear framework exists yet. Malpractice law, FDA oversight, and payer rules were built around human decisions and human-supervised software, not agents that act independently between visits.