JMIR Publications

ABSTRACT

AI-powered therapy chatbots and virtual companions are rapidly entering mental health care. For attention-deficit/hyperactivity disorder (ADHD), they promise timely psychoeducation, behavioral coaching between visits, and structured inputs for measurement-based care across clinics, schools, and homes. Yet practical guardrails—especially for minors—have not kept pace with deployment. This Viewpoint advances a nurse-led, risk-tiered framework to guide safe, effective, and equitable integration of therapy chatbots into ADHD services. Drawing on ADHD care guidelines, emerging AI governance instruments, and lessons from digital health implementation, we identify six persistent gaps: risk classification and predetermined change control plans for adaptive models; child-specific safeguards, including content safety, age gating, and reliable crisis escalation; equity-by-design procurement and reimbursement; interoperable outcomes and safety reporting; workforce readiness with human-in-the-loop oversight led by nursing; and privacy-preserving real-world evidence (RWE). We present a three-layer framework aligning policy levers, implementation enablers, and outcomes, and we compare five complementary policy options: status quo self-regulation; risk-tiered regulation with PCCPs; voluntary certification with public procurement criteria; reimbursement for nurse-delivered digital ADHD care; and a federated RWE and safety surveillance network. A phased roadmap details preparation, sandbox pilots, scale-up with guardrails, and continuous learning, alongside a concise evaluation set spanning access and equity, safety signals, clinical and functional outcomes, experience, and economic value. Taken together, risk-tiered governance, procurement aligned with transparency and age-appropriate design, financing for nurse-led integration, and federated RWE can accelerate safe, equitable adoption of therapy chatbots for youth with ADHD while the evidence base matures.