Lifelong Personalized AI Agents for Executive-Function Training in ADHD: From Childhood to Early Adulthood

Abstract

Attention-Deficit/Hyperactivity Disorder (ADHD) is characterized by persistent impairments in executive functions (EFs), including working memory, inhibitory control, emotional regulation, planning, and sustained attention. These deficits evolve across developmental stages, requiring adaptive and longitudinal interventions tailored to the individual’s cognitive and behavioral profile. Recent advances in artificial intelligence (AI) and personalized learning systems have introduced new possibilities for lifelong support. This paper proposes a conceptual and empirical foundation for Lifelong Personalized AI Agents designed to deliver continuous EF training for individuals with ADHD from childhood to early adulthood. Grounded in theories of neuroplasticity, multimodal learning analytics, and developmental psychology, the study examines how adaptive AI-driven interventions can scaffold executive functioning across school, psychosocial, and early-career contexts. The methodology employs a mixed-methods longitudinal design, integrating digital performance logs, cognitive assessments, and qualitative interviews, analyzed through an AI-augmented modeling pipeline. Early findings from a pilot dataset (N = 86) illustrate significant improvements in working memory, task initiation, and emotional regulation over a 12-month period. This paper lays the groundwork for next-stage development of lifelong EF-support ecosystems informed by personalized AI agents.