The Landscape of Ai-Driven Interactive Storytelling In Adolescent Mental Health: A Scoping Review On Nursing and Communication Perspectives
TL;DR
AI-driven interactive storytelling holds great potential as a complement to decentralized adolescent mental health services, but its optimization strongly depends on improving the AI's linguistic accuracy toward local adolescent language as well as strengthening automatic referral systems to clinical nursing staff to guarantee patient safety.
Abstract
Background: The post-pandemic surge in adolescent mental health disorders calls for innovative and decentralized digital intervention breakthroughs. Artificial Intelligence (AI) technology, through interactive storytelling models and generative chatbots, is now being integrated to provide a safe narrative space for adolescents. However, the computational-linguistic effectiveness and clinical-ethical boundaries of these interventions still require comprehensive mapping. Objective: This study aims to map the landscape, effectiveness, linguistic barriers, and ethical implications of using AI-driven interactive storytelling in adolescent mental health management by integrating psychiatric nursing and digital communication perspectives. Methods: This study used a scoping review design based on the PRISMA 2020 guidelines. Literature was systematically searched across three reputable electronic databases: Scopus, ScienceDirect, and CINAHL. The data selection process used the PCC (Population, Concept, Context) framework strategy. From a total of 146 articles identified in the initial search, 18 final original studies met the inclusion criteria and were extracted for narrative analysis. Results: Synthesis from a digital communication lens shows that adolescents use AI's anonymous venting features to freely express emotional distress. Nevertheless, Natural Language Processing (NLP) still has significant limitations in understanding the dynamics of cyber-slang and crisis metaphors (such as algospeak), which can trigger contextual failures in algorithmic responses. From a psychiatric nursing lens, interactive chatbots have proven effective as instruments for large-scale digital triage and Psychological First Aid. However, the application of this technology must comply with the principle of non-maleficence, in which AI systems must have strict operational limits to detect critical indicators of acute crisis (self-harm and suicidal ideation) and automatically redirect them to professional clinical mental health nurses. Conclusion: AI-driven interactive storytelling holds great potential as a complement to decentralized adolescent mental health services, but its optimization strongly depends on improving the AI's linguistic accuracy toward local adolescent language as well as strengthening automatic referral systems to clinical nursing staff to guarantee patient safety.