30/100

Emerging evidence

Cite
More
EmailLinkedInX

AI chatbots often do worse in children's medicine than adult care, a pediatrics policy statement warns

Study title: Recommendations for the Development and Implementation of Generative Artificial Intelligence Tools in Pediatric Clinical Care: Policy Statement

Evidence type
Commentary, opinion, or narrative review
Sources
1 sourcePeer-reviewed
Published
1 min read

Study title: Recommendations for the Development and Implementation of Generative Artificial Intelligence Tools in Pediatric Clinical Care: Policy Statement

On October 3 2026, Pediatrics published a policy statement on generative AI tools including large language models in pediatric health care. It describes potential uses in decision support, documentation, and education, while noting limited real-world validation and persistent concerns about accuracy, bias, and reliability.

Read the study abstract

The integration of generative artificial intelligence (GenAI) into pediatric health care offers exciting opportunities alongside critical challenges. GenAI tools, like large language models (LLMs), show promise in several health care applications, including but not limited to clinical decision support, documentation, and medical education across a wide range of pediatric subspecialties. However, real-world validation remains limited, and concerns persist around accuracy, bias, reliability, sustainability, and du…

Recommendations for the Development and Implementation of Generative Artificial Intelligence Tools in Pediatric Clinical Care: Policy Statement
Photo by Vitaly Gariev on Unsplash · https://unsplash.com/@silverkblack?utm_source=truvace&utm_medium=referral

The problem

In pediatric care, LLMs often underperform relative to adult specialties and risk exacerbating health disparities due to biased or nonrepresentative training data.

In brief

The statement matters because children and adolescents may face distinct risks from models trained largely on adult data, including poorer performance and widening disparities across race, ethnicity, gender, language, ability, and socioeconomic status. It remains uncertain how to ensure diverse pediatric data, effective oversight, and adaptive governance while preserving trust and privacy.

Main points

  1. Policy statement addresses generative AI and large language models in pediatric clinical care across subspecialties.
  2. Studies cited indicate LLMs often underperform in pediatrics compared with adult specialties, with limited real-world validation.
  3. Recommendations call for diverse pediatric datasets, bias mitigation, privacy safeguards, validation protocols, human oversight, and postmarket surveillance.

The full story

The Pediatrics policy statement published Oct 3 2026 reviews GenAI and LLMs for clinical decision support, documentation, and medical education in pediatrics, noting promise but limited real-world validation.

It flags accuracy, bias, reliability, sustainability, and durability concerns, cites underperformance in pediatrics versus adult specialties, and recommends diverse pediatric datasets, privacy safeguards, rigorous validation, human oversight, pediatric-specific evaluation, postmarket surveillance, and disclosure of GenAI involvement.

Sources

1 on file
  1. Pediatricsdoi:10.1542/peds.2026-079037
    Peer-reviewed2026-10-03

Talk it over

What readers make of this record

What readers say