Artificial Intelligence in U.S. Healthcare: Legal framework, liability, and ethical challenges with special focus on Pediatric Care

Abstract: This paper examines the legal, regulatory, and ethical implications of integrating artificial intelligence into U.S. healthcare, with a focus on pediatric care. It analyzes how ML-SaMD and CDS tools interact with federal oversight mechanisms, including FDA authorization pathways, the 21st Century Cures Act boundary, HIPAA/HITECH privacy requirements, and HTI-1 transparency obligations. The study highlights persistent accountability gaps arising from algorithmic opacity, bias, and uncertainty regarding liability allocation among manufacturers, healthcare institutions, and clinicians. It proposes a responsibility matrix and a pediatric safety framework requiring age-stratified validation, continuous monitoring, explainability adapted to clinical decision-making, and escalation protocols for drift or unsafe behavior. The paper argues that sustainable and equitable AI adoption demands coordinated governance that strengthens patient protection while supporting innovation in systems that directly impact children’s long-term health outcomes. Note: Version updated in repository for preservation and digital accessibility. Original intellectual content remains unchanged. Official reference version:Zenodo DOI: https://doi.org/10.5281/zenodo.17344639Parallel version available:SSRN: https://ssrn.com/abstract=5468846 - DOI: https://doi.org/10.2139/ssrn.5468846Preservation copy at the Internet Archive: https://archive.org/details/zenodo.17344639