Policy

AMA Urges Liability Protections for Physicians Using AI Tools

The medical association wants developers and health systems—not individual clinicians—held responsible when AI makes errors in patient care.

Omega Editorial · November 20, 2026 · 3 min read

Key takeaways

  • The AMA wants AI vendors and healthcare employers—not individual physicians—held liable when mandated AI tools cause patient harm.
  • Roughly 75% of surveyed U.S. health systems had adopted at least one AI tool by February, up from 59% in 2025.
  • The association is calling for laws requiring AI developers to carry liability insurance and disclose product limitations.
  • No AI-related malpractice cases against physicians have been filed yet, but legal experts say such litigation is likely coming.
  • The AMA framework states doctors should be able to opt out of using AI tools without penalty when clinical validation is uncertain.

The American Medical Association is pushing lawmakers to establish clear liability protections for physicians who use artificial intelligence in clinical settings, arguing that individual doctors should not face legal consequences when AI tools they're required to use produce harmful errors.

On November 10, the AMA called on state and federal legislators to ensure that AI vendors, manufacturers, and healthcare employers—not frontline clinicians—bear responsibility when autonomous systems cause patient harm. The move comes as AI adoption accelerates across U.S. healthcare, with approximately 75% of 120 surveyed health systems having deployed at least one AI tool as of February, according to healthcare research firm Eliciting Insights. That figure represents a jump from 59% adoption in 2025.

Why it matters

As hospitals rapidly implement AI scribes, diagnostic assistants, and administrative tools, the legal framework governing medical errors has not kept pace. Physicians increasingly find themselves caught between institutional mandates to use AI and personal liability for outcomes they cannot fully control. Without clear statutory protections, the first wave of AI-related malpractice cases could create precedent that either chills beneficial innovation or unfairly punishes individual practitioners for systemic technology failures.

The liability gap

The AMA's position reflects growing concern that existing malpractice frameworks don't account for AI's unique characteristics. While many AI systems operate with a human-in-the-loop—a clinician who reviews and validates AI output—doctors cannot catch every error, particularly as these tools become more autonomous.

"AI can help physicians analyze information, identify patterns, and support clinical decision-making, but patients deserve to know who is responsible for their care," AMA President Willie Underwood III said in a statement.

The association is also advocating for requirements that AI developers carry liability insurance and provide transparent data on product limitations and performance metrics. The AMA emphasized that contracts, patient consent forms, and disclosure documents should not be used to shield developers from accountability.

Precedent and practice

Legal experts told Modern Healthcare that courts have established precedent for holding AI developers—rather than end users—responsible for tool failures, though healthcare-specific case law remains limited. Healthcare Dive reported that while no AI-related malpractice cases against physicians have been filed yet, such litigation appears inevitable as deployment expands.

Sharif Vakili, a primary care physician and CEO of telehealth company UpDoc, drew parallels to existing medical devices like implantable defibrillators. In those cases, physicians are responsible for device selection and configuration, while manufacturers ensure design safety and performance.

"Accountability should follow authority and control," Vakili said. "Developers should be responsible for the safety and performance of their systems, health systems for how they deploy and monitor them, and clinicians for the decisions they actually make."

The AMA released a separate framework on October 29 stating that physicians should be able to decline AI tools without penalty when clinical validation or standard-of-care status remains uncertain.

Jan Carney, president of the American College of Physicians, told Healthcare Brew that her organization shares the AMA's concerns, emphasizing that AI should support rather than complicate the patient-physician relationship.

These details were first reported by Healthcare Brew.

  • #medical ai
  • #healthcare liability
  • #physician protection
  • #medical malpractice
  • #healthcare regulation
  • #clinical ai

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy