Artificial Intelligence - Enabled Mobile Exercise Prescription: Modelling, Explainability, and Health System Integration (Preprint)

UNSTRUCTURED Physical inactivity remains a major contributor to global morbidity and mortality and represents a substantial burden for healthcare systems worldwide. Mobile health technologies and wearable devices have emerged as scalable tools to promote physical activity; however, many existing digital interventions rely on static goal setting and have demonstrated limited long-term engagement. Advances in artificial intelligence (AI) are enabling a transition toward adaptive mobile exercise prescription systems that leverage high-frequency behavioural and physiological data to deliver personalised activity recommendations. In this editorial, we discuss the emerging role of AI-enabled mobile exercise platforms, focusing on advances in behavioural modelling, reinforcement learning–based recommendation systems, and just-in-time adaptive interventions. We also examine key challenges related to generalisability, algorithmic transparency, human factors, and digital equity. Finally, we consider the requirements for integrating AI-driven exercise prescription into healthcare systems, including clinical validation, regulatory oversight, and alignment with digital health evaluation frameworks. AI-enabled exercise prescription has the potential to deliver personalised preventive interventions at scale; however, its impact will depend on methodological rigour, transparency, and responsible integration within healthcare pathways.