As physicians and their healthcare organizations increasingly adopt augmented intelligence (AI) into the everyday practice of medicine—including sometimes incorporating it into decision making of how to proceed with a patient’s treatment plan—ethical questions about how to use the technology arise.
These questions include:
- Can AI replace clinical decision-making?
- What if a patient asks me to use AI for a second opinion or to make their medical decisions?
- What if a patient asks me not to use AI during their medical treatment?
- How much detail do I actually have to put in my note regarding my use of AI?
- How do I know if I am relying on the AI tool too much?
The AMA recently introduced the “Ethical AI Use in Medicine Series,” a new CME program that helps physicians navigate these and other questions to safely, effectively and ethically use AI in clinical practice.
“As AI rapidly transforms medical practice, the AMA is committed to equipping physicians with clear, practical ethics guidance rooted in our Code of Medical Ethics,” said AMA President Willie Underwood III, MD, MSc, MPH. “By translating these ethical principles into actionable steps, we aim to support physicians in using AI tools responsibly—always with patient trust and safety at the forefront. These resources empower physicians to navigate new technology with confidence, reinforcing the AMA’s role as a trusted source for clinical excellence and ethical standards.”
The AMA collaborated with the Duke Institute for Health Innovation and Health AI Partnership to create the series, one element of which is “Ethical AI Use in Medicine: A Clinical Practice Toolkit.”
The toolkit outlines three pillars of responsible AI use in clinical practice and lays out a chart that describes what typical AI tools fall into a low, moderate or extensive oversight category.
Further, the toolkit outlines a six-step process so that physicians and their practices can:
- Evaluate tool appropriateness.
- Evaluate ethical considerations and evaluate health outcome impacts.
- Evaluate ethical considerations when it comes to transparency and disclosure.
- Evaluate ethical considerations when it comes to transparency and permission and informed consent.
- Evaluate ethical considerations when it comes to accountability and oversight, including who is responsible.
- Incorporate continuous post-implementation evaluation and ongoing monitoring.
The toolkit also includes a case study that demonstrates how to apply the toolkit.
“AI is rapidly reshaping clinical practice, making ethical use, patient-centered communication and ongoing evaluation essential,” the toolkit says.
Other modules in the “Ethical AI Use in Medicine Series” are:
- “Ethical AI Use in Clinical Decision-Making,” which engages learners in realistic clinical scenarios that strengthen judgement, accountability, and patient-centered decision-making when AI informs care.
- “A Structured Decision Simulator for Using AI Mortality Prediction for Clinical Decision Support: AI Ethics in Clinical Practice,” which immerses physicians in a guided decision simulator focused on the responsible use of an AI-powered mortality prediction model, including risk communication, bias awareness and appropriate clinical reliance.
- “A Structured Decision Simulator for Using an Ambient AI Scribe: AI Ethics in Clinical Practice,” which helps physicians and other health professionals apply ethical safeguards to ambient documentation, with emphasis on transparency, patient trust, physician verification and safe use of AI-generated notes.
The AMA says that “AI should be viewed as an augmenting tool that complements human judgment, reinforces evidenced-based practice and supports clinicians in delivering high-quality, patient-centered care.”
As AI continues to evolve, the AMA will, among other things, foster evidence-based medicine in AI tools, collaborate with medical specialty societies, relevant key parties, regulators and AI developers to establish standards and develop a framework for evidence attribution, evaluation and validation in AI clinical decision support systems.
From AI implementation to digital health adoption and EHR usability, the AMA is fighting to make technology work for physicians, ensuring that it is an asset to doctors. That includes the AMA Center for Digital Health and AI, which works to ensure physicians help shape how AI is developed, implemented and regulated across the healthcare system, with patient safety and physician-led care remaining at the center of those efforts.
Bringing medical ethics to life
Also recently introduced is the “AMA Ethics Pearls in Practice Series,” a set of case-based education modules from the AMA Code of Medical Ethics that offers practical resources to help physicians and healthcare professionals apply the code in their daily work. The series offers concrete advice on how health professionals should interact with patients and address critical decisions within their practice, including AI. The modules also provide links to relevant AMA Code of Medical Ethics opinions.
“The code doesn’t do doctors and the patients they care for any good if it just sits on the wooden or digital shelf. It needs to be understood and applied,” said Audiey C. Kao, MD, the AMA’s vice president for ethics. The new modules “help to situate the code and its ethical guidance in practical settings and dilemmas that physicians may encounter during their clinical careers, during their educational careers and so forth.”
In “What Are the Ethical Responsibilities of Using AI in a Clinical Setting?” physicians are presented with a case of a patient with atrial fibrillation and whether the use of AI-assisted robotic thrombectomy is appropriate in this circumstance.
From this case-based springboard, the module explores key ethical pearls to help physicians understand the types of AI tools, evaluate tool appropriateness, address individual ethical scenarios, disclose AI use, obtain informed consent and maintain oversite so that physicians can foster trust and make responsible, collaborative decisions that uphold the integrity of the patient-physician relationship.
For example, the scenario prompts the question of whether AI can replace clinical decision-making. In one of the ethical pearls, the module explains that “AI augments, but does not replace clinical decision making or physician ethical responsibility” and that “replacing physicians with AI or solely depending on AI for decision-making would undermine the patient’s trust essential to a therapeutic patient-physician relationship.” And it explains that physicians should critically evaluate all AI outputs and that AI should be:
- Supplemented with additional clinical context and physician judgement.
- Adapted to patient-specific needs and values.
- Overridden when outputs conflict with sound clinical judgment or patient goals.
The module further offers links to continued learning through relevant AMA Code of Medical Ethics opinions, including ones for electronic communication with patients and ethically sounds innovation in medical practice.
The CME module, “What Are the Ethical Responsibilities of Using AI in a Clinical Setting?” is enduring material and designated by the AMA for a maximum of 0.25 AMA PRA Category 1 Credit™.
It is part of the AMA Ed Hub, an online learning platform that brings together high-quality CME, maintenance of certification, and educational content from trusted sources, all in one place—with activities relevant to you, automated credit tracking, and reporting for some states and specialty boards.
Learn more about AMA CME accreditation.
