The American Medical Association is calling for laws that would protect doctors from liability when AI harms patients, even if they used the tool as intended. Its proposal would shift more responsibility towards AI companies and vendors, and in some cases health systems.
Watch Desk analysis
What happened
The AMA is pressing for federal and state legislation covering AI used in healthcare, Axios reports. The group also wants AI companies to carry adequate liability insurance and disclose product limitations and performance data.
Under the AMA’s proposed approach, hospitals, health systems, insurers and health technology vendors could be responsible when required use of a specific AI product contributes to harmful errors. These are the group’s policy demands, not a description of laws already in force.
Why it matters
AI tools can influence decisions such as whether to order a test or prescribe a costly drug. When a system’s inputs or limitations are unclear to the doctor using it, deciding who should answer for harm becomes more complicated than pointing at the person holding the keyboard.
The proposal puts two questions together: who carries the legal risk, and who must disclose how a system performs? Shifting responsibility towards suppliers could recognise that clinicians cannot reasonably check every opaque tool. It could also leave difficult questions about how fault is divided when a doctor, hospital and vendor all play a part.
Our read
The AMA’s case is strongest where clinicians are required to use a tool but cannot see enough to check its recommendation. That does not settle where responsibility should land in every case. A workable policy would need to distinguish a supplier’s product failure from a clinical decision, without treating “the AI did it” as either a defence or a complete explanation.
What to watch
- Whether federal or state lawmakers take up the AMA’s proposals.
- What the group means by adequate insurance and required performance disclosures.
- How regulators address monitoring of AI-enabled medical devices after deployment.
Discussion spark: If a healthcare provider requires doctors to use an AI tool, should the provider share liability for its errors, or should responsibility sit mainly with the tool’s maker?
Sources and evidence
- Doctors want liability shield from AI mistakes (13 November 2026, 10:30 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.