YOUR GO-TO INTELLIGENCE RESOURCE FOR THE SCHOLARLY PUBLISHING INDUSTRY

The American Medical Association (AMA) has introduced a comprehensive policy framework to protect physicians from unauthorized use of AI‑generated deepfakes. The framework, developed by the AMA Center for Digital Health and AI, aims to modernize identity protections and address legal gaps to safeguard patient safety, professional integrity, and public trust.

AMA CEO John Whyte, emphasized that AI deepfakes impersonating physicians represent a public health and safety crisis. He explained that when malicious actors exploit a doctor’s identity, they erode patient trust and may mislead individuals toward harmful or unproven treatments. He urged federal and state lawmakers to take strong action to protect physician identities, ensure transparency, and prevent fraud, noting that safeguarding professional integrity is essential to maintaining trust and delivering quality care in a rapidly evolving digital environment.

Deepfakes have already been used to impersonate physicians, manipulate public opinion, and endorse unproven treatments, creating risks for both individual patients and the broader health care system. Such impersonation undermines the patient‑physician relationship, weakens confidence in evidence‑based care, and exposes the public to deception.

The AMA framework is built around seven policy principles:

• Physician identity protections: Names, images, likenesses, voices, and digital replicas are protected and may only be used with explicit, informed consent.

• Ban on deceptive impersonation: AI‑generated or altered content impersonating physicians without consent must be prohibited and treated as deception.

• Opt‑in, revocable consent: Physician identity use requires explicit, separate consent specifying purpose, audience, and duration, with the option to revoke.

• Mandatory labeling and transparency: AI‑generated depictions must be clearly labeled, watermarked, and disclosed to patients before interaction.

• Shared responsibility: Platforms, hospitals, and vendors must implement safeguards, including rapid takedown mechanisms and clear labeling.

• Enforcement and remedies: Physicians must have access to procedures for documenting misuse, triggering takedowns, and pursuing remedies, with federal agencies empowered to enforce laws.

• Minimizing administrative burden: Protections should be the default, with standardized consent processes that do not impose undue burdens on physicians.

The AMA stated that this framework will guide its collaboration with lawmakers, regulators, and industry partners to prevent AI‑generated deepfakes of physicians and protect both patients and medical professionals.