Introduction: Preterm infants face risks of growth retardation and readmission after discharge, yet traditional follow-up models suffer from poor compliance and insufficient multi-sector coordination. Objective: To develop a hospital-community-family artificial intelligence (AI) follow-up system for preterm infants and evaluate its effectiveness. Methods: A questionnaire based on the Capability–Opportunity–Motivation–Behavior (COM-B) model was first administered to 202 parents of preterm infants from three tertiary hospitals (Sept–Nov 2024) to analyze follow-up behaviors and influencing factors. Based on these findings, an AI-driven follow-up system was developed. Subsequently , 146 preterm infants discharged from Shanxi Bethune Hospital (Dec 2024–May 2025) were randomly assigned to a control group (conventional follow-up, n=73) or an intervention group (AI system application, n=73). Parental knowledge, parenting stress, infant follow-up rates, growth parameters, Gesell developmental scores, and 6-month readmission rates were compared. Results: The AI system comprised three modules (capacity building, communication/support, motivation enhancement) with 12 functions. Parents in the intervention group demonstrated significantly higher knowledge and skill scores and lower parenting stress than those in the control group (P < 0.05). Infants in the intervention group showed better growth outcomes, higher neurobehavioral scores at corrected age of 6 months, and lower readmission rates (P < 0.05). Conclusion: The hospital–community–family integrated AI follow-up system improves parental competence and infant developmental outcomes, providing an effective model for continuity of care in preterm infants.