To the Editor We read with great interest the Viewpoint by Gazit and colleagues proposing that artificial intelligence (AI) in critical care should preferentially support de-escalation rather than escalation of care. We concur with the authors’ perspective and would like to further emphasize additional reasons why de-escalation represents the phase in which AI may deliver its greatest clinical value.
JAMA Pediatrics published a clinical update in Research Highlights on 06 Apr 2026.
The item focuses on Artificial Intelligence and De-Escalation of Critical Care.
Review the original article for the full source wording and details.