Artificial Intelligence for Weight Management in Children: A Narrative Review

Background/Objectives: Childhood overweight and obesity represent a major global public health challenge, with increasing prevalence and significant long-term metabolic, cardiovascular, and psychosocial consequences. Standard pediatric weight-management strategies based on lifestyle modification often achieve modest and variable results, highlighting the need for more personalized and scalable approaches. Artificial intelligence (AI) has emerged as a promising tool to enhance prevention, early risk stratification, and management of pediatric overweight and obesity. Methods: This narrative review was conducted through a structured search of PubMed, Scopus, and Web of Science for English-language studies published up to January 2026. The main search terms included “artificial intelligence”, “machine learning”, and “deep learning”, combined with “child”, “adolescent”, “pediatric”, “childhood obesity”, “pediatric overweight”, “body mass index”, “weight management”, “nutrition”, “diet”, “physical activity”, “lifestyle”, and “behavior change”. After title/abstract and full-text screening according to predefined eligibility criteria, the included studies were qualitatively synthesized and grouped by main application domains. The initial database search identified 412 records. After removal of 96 duplicates, 316 records were screened by title and abstract. Full-text assessment was subsequently performed for 175 potentially eligible articles. Following this evaluation, 51 studies met the eligibility criteria and were retained from the database search. Additional relevant articles were identified through manual screening of reference lists and related reviews, resulting in the final set of studies included in the narrative synthesis. Results: The review identified five main domains of AI application in pediatric weight management: risk assessment and prediction, dietary assessment and nutritional support, physical activity and lifestyle monitoring, behavioral and psychological support, and clinical decision support. Across the included literature, AI-based approaches were most frequently applied to predictive modeling using longitudinal BMI or growth trajectories, birth characteristics, parental BMI, sleep duration, physical activity, sedentary behavior, and family or socioeconomic factors. However, the evidence base was largely composed of observational and predictive-modeling studies, whereas interventional studies, real-world implementation studies, and long-term pediatric weight-outcome data remained limited. Conclusions: This narrative review indicates that AI has potential as a complementary tool within multidisciplinary, family-centered pediatric weight-management pathways, particularly for early risk stratification, personalized monitoring, and behavioral support. However, the findings also highlight that current evidence remains mainly exploratory and predictive rather than interventional. Further longitudinal, real-world, and ethically grounded research is required to confirm effectiveness, safety, clinical usefulness, and equitable implementation in pediatric populations.