Related Experiment Videos
Building the Foundations for Trustworthy AI in Pediatrics
Samuel G Finlayson1,2,3, Aaron Wightman1,2,4,5, Elliott Mark Weiss1,2,4,6
1Seattle Children's Hospital, Seattle, Washington.
Pediatrics
|July 20, 2026
Summary
Implementing trustworthy artificial intelligence (AI) in pediatrics requires addressing unique challenges in data, development, and infrastructure. Foundational systems and governance are crucial for beneficial AI integration in child healthcare.
Area of Science:
- Pediatric Health
- Medical Artificial Intelligence
- Health Informatics
Background:
- Artificial intelligence (AI) adoption in healthcare is accelerating, with significant potential for pediatrics.
- Existing ethical frameworks for trustworthy AI in pediatrics exist, but foundational infrastructure needs are underexplored.
Purpose of the Study:
- To examine the unique challenges at the intersection of AI and pediatric care.
- To outline the necessary infrastructure for trustworthy AI in pediatrics.
- To provide recommendations for stakeholders.
Main Methods:
- Analysis of challenges posed by AI in pediatrics (e.g., errors, cognition, infrastructure).
- Analysis of challenges posed by pediatrics for AI (e.g., data, developmental heterogeneity, patient autonomy).
- Outline of essential components: data systems, governance, validation, and public trust infrastructure.
Main Results:
- AI presents unique risks in pediatrics, including non-intuitive errors and impacts on clinicians.
- Pediatrics poses unique challenges for AI development, such as data scarcity and developmental variability.
- Establishing robust data systems, governance, validation, and public trust is essential.
Conclusions:
- Foundational infrastructure is critical for safe and effective AI implementation in pediatrics.
- Collaborative efforts among clinicians, policymakers, and technologists are needed.
- Recommendations are provided to guide the development of trustworthy AI in child healthcare.