By: Annabelle Racine, Intern, National Health Council  

Artificial intelligence (AI) is quickly becoming part of everyday health care. It is being integrated into tasks from drafting clinical notes and responding to patient portal messages to supporting diagnosis and treatment decisions. These tools aim to improve efficiency and patient care. As these technologies become more common and advanced, one question remains: How should patients and caregivers be informed when AI is involved in care?  

During my internship with the National Health Council (NHC), I had the opportunity to explore how thoughtful AI disclosure can help ensure that innovation strengthens trust, supports communication, and keeps patients informed as health care continues to evolve. Through a literature review and gap analysis conducted over the summer of 2026, I examined how AI disclosure frameworks and communication practices are evolving in AI-enabled health care. This work included reviewing the ethical foundations of informed consent, examining the current applications of AI in health care, identifying key gaps in existing guidance, and developing patient-centered recommendations to help inform future work by the NHC and the Patient Experience + Innovation (PXI) Center. 
 
Informed consent supports patient understanding, autonomy, and shared decision making through communication between patients and clinicians. Although many AI applications raise questions about disclosure rather than formal consent, these principles remain relevant to how patients are informed when AI influences their care. 

Overall, the literature suggests that patients want the health care system to be transparent about its use of AI: They want meaningful information about AI’s role in their care, reassurance that clinicians will remain responsible for clinical decisions, and communication that supports informed health care decisions. Meaningful disclosure requires information patients can understand and use in decisions about their care. 
 
However, when disclosing AI use to patients, not all applications require the same level of disclosure. While AI used for basic administrative tasks, such as scheduling and billing, may call for less detailed disclosure, AI used for diagnosis, treatment, or patient communication may require more detailed disclosure. A risk-based approach can promote flexibility and reflect the varying impact AI can have on patient care.  

Although professional organizations have begun releasing guidance for responsible AI implementation, which have established strong foundational principles, gaps remain. Existing recommendations emphasize transparency and human oversight with AI but provide limited practical detail about what information patients want to receive, how communication should vary across AI applications, or how communication should be adapted to patient preferences and health literacy. Another limitation of many of these guidelines is that patient and caregiver perspectives remain underrepresented in their development, which largely reflects the perspectives of health care and professional organizations. Additional patient and caregiver input could help clarify how information needs may differ by health literacy, age, culture, or previous experience with AI.  

The goal for the future of AI in health care should be that patients remain informed, empowered, and at the center of every health care decision. This presents an important opportunity: for the NHC and the PXI Center: Together with patients, caregivers, health care organizations, and technology developers, the NHC can play an important role in creating patient-centered approaches to AI disclosure that build trust and support innovation. Potential work could include gathering patient and caregiver input on disclosure preferences and developing practical guidance on when AI should be disclosed, who should communicate its use, and how communication should reflect AI’s role and potential impact on care. 

Interested in driving the future of patient-centered AI? Get involved with the PXI Center to help ensure patient and caregiver perspectives shape how AI is developed and used in health care. Contact Spencer Morrissey, Director, PXI, for more information.  
 
On Oct. 7, from 2 to 4 p.m. ET, join the Q3 PXI Convening, “Health AI Regulation × Patient Protection.” The program features a presentation from NHC member Aimed Alliance on its 2026 report on patient protections, followed by a panel exploring state, federal, and global perspectives on AI regulation. Register to attend.