Pediatricians now have formal guidance for putting generative AI to work in the clinic. The American Academy of Pediatrics warns that tools built for adults do not serve children well, so the specialty needs its own rules rather than borrowed ones.
The academy spelled out the advice in a policy statement whose formal title runs: “Recommendations for the Development and Implementation of Generative Artificial Intelligence Tools in Pediatric Clinical Care.” The AAP’s journal, Pediatrics, printed it in the November 2026 issue. San Diego hosts the organization’s National Conference and Exhibition, where the document is set for presentation on October 3, the same day it appeared online.
Lead author Srinivasan Suresh sees real gains in decision support, documentation, education and the everyday clinical workflow. He also argues the field must move into this territory with what he calls rigorous human oversight and accountability.
Closing that gap is the point. Children appear far less often than adults in the data that trains these tools, and few are tested on young patients in particular. Developers, the academy says, should handle child data responsibly, confront bias, and design privacy protections matched to how children develop and what families need.
The guidance also presses for tools that reflect the range of races, ethnicities, languages, abilities and incomes found among families, so AI does not deepen health disparities. It stays within clinical care, leaving direct use by children and families to separate advice the academy is still preparing.
