The American Medical Association adopted policy on Monday seeking legal protections for physicians who decline artificial intelligence tools or recommendations in clinical care.
Meeting at its Interim Meeting near Orlando, Florida, the association’s House of Delegates resolved to push for laws placing product liability for defective clinical AI on vendors rather than doctors. The policy positions are advocacy goals.
Existing malpractice rules stay unchanged in every state. Core debates regarding physician accountability, medical oversight, and patient care obligations arise as artificial intelligence takes on clinical duties.
The American Medical Association and the Federation of State Medical Boards convened more than 200 physicians, medical board representatives, patients, policymakers, and technology innovators on September 14, 2026. The event united diverse viewpoints.
These insights inform the AMA Center for Digital Health and AI across governance, economic sustainability, and care transformation. The center collaborates with stakeholders throughout healthcare, government, technology sectors, and patient advocacy groups.
Malpractice law generally asks whether a physician met the standard of care. Legal scholars warn that doctors could face blame for overriding a correct AI recommendation and for following a flawed one.
Data from a Journal of Nuclear Medicine research project involving 2,000 American adults simulating jury members showed that evaluations of medical negligence were influenced by whether practitioners chose to follow or dismiss AI guidance.
AI Cannot Hold Physician Licensure or Professional Titles
The newly adopted guidance establishes that artificial intelligence can assist with clinical tasks but cannot hold physician licensure, professional titles, or certification. The association will advocate for model legislation that bars presenting nonhuman entities with titles reserved for physicians.
Technology can support medical professionals, yet artificial intelligence is entirely ineligible to serve as a physician, according to the updated guidelines and an AMA report issued on November 10, 2026.
“Patients deserve to know who is responsible for their care,” said AMA President Dr. Willie Underwood III, according to the association’s meeting highlights. Medical practitioners must be granted proper authority, supervision, and legal shields to exercise autonomous judgment whenever artificial intelligence participates in patient treatment, he said.
Physicians who use AI tools as intended and exercise appropriate clinical judgment but the product still causes harm should receive liability protections, the association argues.
The Food and Drug Administration had authorized more than 1,450 AI-enabled medical devices as of December 2025. Specialty mattered too, with oncology and hematology (2.00), internal medicine (1.88), and family medicine (1.80) running above their workforce share.
Developers Must Maintain Liability Coverage and Disclose Metrics
The framework advocates for statutory requirements mandating that technology developers maintain sufficient liability coverage. Developers must reveal performance metrics and technical constraints.
Developers must also refrain from using contractual clauses that transfer liability for software malfunctions onto medical providers or hospital networks.
The new policy builds on a clinical AI liability framework the AMA released in October. Adoption rates among practitioners continue to rise.
In an AMA survey of nearly 1,700 physicians released in March, 81 percent reported using artificial intelligence in their practices, more than double the share recorded in 2023.
It is unclear how state legislatures will translate these advocacy goals into binding statutes.