AMA Seeks Liability Protections for Doctors Who Decline Clinical AI Tools

The American Medical Association is seeking federal and state laws to protect doctors who decline clinical artificial intelligence tools or reject their recommendations, while directing liability for defective products toward the companies behind them.

Delegates approved the measure during the association’s interim meeting in Orlando on Monday, Nov. 9. It calls for protection when a physician chooses not to use an AI tool or not to follow its recommendation. For harm caused by a defective product, the policy would assign responsibility to a vendor, manufacturer or distributor if the physician used the tool as intended and exercised appropriate clinical judgment.

“Patients deserve to know who is responsible for their care,” AMA President Willie Underwood III said.

The policy also covers situations in which a hospital, health system or insurer requires doctors to use a specific AI product. If errors from that product cause harm, the policy would assign responsibility to the organization that required its use and the product’s vendor.

The AMA is calling for legislation that would require vendors and manufacturers to carry adequate liability insurance and disclose known failure rates, product limitations and performance data. It also opposes contract terms that transfer responsibility for AI product failures to physicians or health systems.

An earlier AMA framework allocates responsibility among developers, health care organizations and physicians based on their authority and ability to prevent harm. It also preserves avenues for patients to seek accountability and remedies. The new policy places AI within physician-led care and says AI should not receive physician licensure, professional titles or certification to practice medicine.

The AMA plans to pursue legislation to establish the proposed protections. Adoption of the policy does not itself change the law or create new liability rules.