Authors Yuan Shen and Linghua Yu have published a commentary in Nature arguing that the governance of artificial intelligence in healthcare must prioritize patients rather than focusing solely on institutional stakeholders.
The critique targets a framework proposed by Kyle Lam and colleagues, which was published earlier in the journal. That framework categorizes medical AI based on autonomy, automation, and operational scope to establish regulatory and liability boundaries across different capability levels. While the authors of that study acknowledged the importance of informed consent, Shen and Yu contend that their model remains predominantly oriented toward distributing responsibility among clinicians, healthcare organizations, device manufacturers, and regulators.
According to the new correspondence, patient rights are not systematically integrated into the design of such governance structures. The authors suggest that a more balanced approach is necessary to ensure that the voices and rights of patients are adequately protected as AI becomes increasingly embedded in medical practice.
The piece appears in volume 657 of Nature, published on September 8, 2026.
Informed consent alone doesn’t cut it. We need patients at the table during AI design, not just after deployment.
But who defines patient rights? This sounds noble, yet implementing a patient-centric model without clear metrics seems nearly impossible.
Finally, someone is highlighting the gap in current frameworks. Patients shouldn’t just be passive recipients of care.