As AI becomes increasingly integrated into clinical care, the American Medical Association House of Delegates has adopted new policies to ensure AI remains under physician oversight and does not replace physician decision-making.
Adopted by physicians and medical students at the AMA House of Delegates' annual meeting, the policies focus on AI use in clinical decision support and in health insurance coverage determinations.
AI-based coverage decision policy
First, the House of Delegates adopted a policy to create additional safeguards for AI-based coverage decisions, ensuring they are based on up-to-date, evidence-based medical information and reviewed by physicians. The policy also opposes the use of autonomous or semiautonomous AI technologies in place of physician review in coverage determinations.
Physicians have previously expressed concerns about AI utilization in prior authorization, with a 2025 AMA survey showing that 61% of physicians were alarmed by healthcare payers' increasing reliance on AI in this area.
The policy adoption also comes as Senate Democrats introduced a resolution to roll back a pilot of the Wasteful and Inappropriate Service Reduction (WISeR) Model, which aims to explore approaches to expedite the prior authorization process within Medicare using AI.
The AMA plans to advocate for greater transparency in AI-driven prior authorization and other utilization management decisions and for regulations requiring AI-enabled tools to be integrated into physician-led processes.
"When health plans use AI-driven tools to deny or delay care without explaining how those decisions were reached, physicians and patients are left in the dark," said AMA CEO John Whyte, M.D., in the press release. "AI should never function as an unaccountable black box. Health plans must be transparent about how these tools work, what evidence and data sources they rely on, and whether a qualified physician reviewed the decision."
AI-based clinical decision support policy
The AMA also adopted a policy to ensure that AI-based clinical decision support tools reflect the principles of evidence-based medicine. The association will work with medical specialty societies, regulators, AI developers and other stakeholders to create standards for evidence attribution, evaluation, validation, transparency and explainability in AI clinical decision support tools.
Further, the AMA will advocate for regular audits of AI-driven clinical review tools. These audits may be triggered by significant changes to AI models, training data or clinical guidelines, as well as comprehensive annual reviews.
"AI has enormous potential in healthcare, but it cannot replace physician judgment," Whyte said. "Patients deserve care decisions that are informed by the latest medical evidence and guided by a physician who understands their individual needs. Whether AI is helping a physician make a clinical decision or assisting with an insurance review, there must always be transparency, accountability, and meaningful physician oversight."
Physicians' concerns around AI-based clinical decision-making are backed by research. Recent research shows that large language models struggle with some clinical reasoning tasks, and while advanced models can outperform physicians on other tasks, they cannot autonomously practice medicine.
Anuja Vaidya has covered the healthcare industry since 2012. She currently covers healthcare IT and innovation, including artificial intelligence, digital healthcare, EHRs and interoperability.
