Large Language Models in the Assessment and Care of Internet Gaming Disorder

Internet Gaming Disorder (IGD), recognized in International Classification of Diseases (ICD-11), affects millions—especially adolescents and young adults—and poses chal-lenges that invite scalable innovations in care. This narrative review examined how Large Language Models (LLMs) could support IGD prevention, assessment, treatment, and research. We conducted targeted searches of PubMed, Scopus, Google Scholar, and IEEE Xplore for 2010–October 2025, supplemented by backward/forward citation chasing; English, peer-reviewed clinical, methodological, and review work was priori-tized; as a narrative review, we did not apply PRISMA or perform quantitative syn-thesis; in total, we synthesized over 50 sources. We synthesize peer‑reviewed, IGD‑specific AI/ML studies with explicit reporting of training approach, valida-tion/performance, dataset size, and model openness. Preliminary improvements ob-served in adjacent digital‑health trials highlight promise yet underscore the need for rigorous, IGD‑specific validation; to date, IGD‑specific randomized trials remain scarce. Evidence spans transformer‑embedding text screening with supervised regres-sion (r ≈ 0.48), multimodal EEG + neuropsychology classification (≈71% vs. comparison groups), fNIRS deep learning (≈88% vs. healthy), and fMRI‑based connec-tomics/MVPA, with sample sizes n = 40–417 and most implementations being re-search‑only (no public code/data). Principal concerns include privacy and data gov-ernance, algorithmic bias, inconsistent crisis-escalation performance, and a nascent clinical evidence base. We conclude that LLMs may augment—but should not re-place—human clinicians; near-term promise lies in hybrid human-AI pathways, mul-timodal integrations with wearables and gaming APIs, and rigorous prospective trials to establish safety, effectiveness, and equity in IGD care.