Key takeaways:
- At its annual meeting, the AMA adopted new policies stressing the need for physician oversight.
- They also advocate for more transparency when AI is used in both clinical and in health insurance decision-making.
The AMA adopted new AI policies in a bid to ensure that the technology supports evidence-based medicine, bolsters patient care and serves under a physician’s oversight, instead of replacing their discernment.
According to an AMA press release, AI technologies could help with efficiency and synthesizing information, but there are still important concerns for bias, long-term impact on both physicians and patient outcomes, explainability and transparency.
At its annual House of Delegates meeting, the AMA adopted new AI-related policies. Image: Adobe Stock
These new policies, which address AI’s growing popularity in both the clinical and health insurance decision-making realms, emphasize that AI should only be used as an assistive tool, rather than “an autonomous decision-maker,” according to the release. The policies also call for accountability, transparency and oversight from physicians every time AI is used in patient care.
“AI has enormous potential in health care, but it cannot replace physician judgment,” John Whyte, MD, MPH, CEO of the AMA, said in the release. “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. Technology should support better care — not stand between patients and the care they need.”
The AMA has said it will work with key stakeholders — including regulators, medical specialty societies and AI developers — to create standards for evidence transparency, evaluation, attribution, validation and explainability in systems that support clinical decision-making. This is in the hope of ensuring AI tools “reflect the principles of evidence-based medicine and provide physicians with information they can understand, evaluate and trust,” according to the release.
Another AMA policy calls for regulations to guarantee that health coverage decisions — which are increasingly based on AI, according to the release — are reviewed by physicians in appropriate fields with a foundation of evidence-based, up-to-date medical information.
The policy additionally called for safeguards requiring AI technologies be integrated into a physician-led process and increased transparency when AI is involved in prior authorization decisions, including the disclosure of any guidelines, data sources or clinical logic used in adverse decisions. In this vein, the AMA is advocating for regular audits of clinical review tools driven by AI to help strengthen accountability. That also includes audits triggered by “significant changes to” training data, clinical guidelines or AI models themselves, “as well as comprehensive annual reviews to ensure continued alignment with standards of care,” according to the release.
“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,” Whyte said in the 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.”
Perspective
The idea that AI should “support, not replace” physician judgment is not controversial. It’s obvious. The real issue is that, across health systems today, we are already drifting in the opposite direction — quietly, incrementally and often without enough oversight. AI tools are being embedded into workflows, influencing decisions and shaping clinical behavior faster than governance models can keep up. How many physicians are actually aware that they are using AI?
The greatest risk is not that AI will replace physicians overnight. It’s that it will subtly erode clinical judgment through automation bias, opaque recommendations and overreliance on systems that clinicians neither fully understand nor control. In that sense, the danger is not disruption, it’s complacency.
What the AMA policy stops short of saying explicitly is this: physician oversight alone is not enough. Oversight without infrastructure, governance and accountability is performative. Health systems need enterprise-level AI operating models — clear standards for validation, monitoring and lifecycle management — because fragmented, point-solution AI adoption is already creating measurable clinical and operational risk.
Deepti Pandita, MD, FACP, FAMIA
University of California Irvine Health
Disclosures: Pandita reports no relevant financial disclosures.
Ask a clinical question and tap into Healio AI’s knowledge base.
- PubMed, enrolling/recruiting trials, guidelines
- Clinical Guidance, Healio CME, FDA news
- Healio’s exclusive daily news coverage of clinical data
