From Support to Maladaptive Reliance: A Scoping Review of the Psychosocial Roles, Risks, and Boundaries of AI Chatbot Use Among Young Adults

Abstract Background In recent years, developments in artificial intelligence have resulted in chatbots transitioning from basic information providers to being used in more advanced and complex forms of conversations with humans. From the findings of previous studies, they indicated that individuals use AI chatbots not only for the purpose of obtaining information or answers, but also for guidance and emotional support. Human beings seem to feel more comfortable in expressing their thoughts and emotions when they recognize an environment as nonjudgmental. This may lead them to the usage of AI chatbots for seeking emotional and social support. Thus, it is crucial to map and analyze how these social experiences may influence individuals, especially young adults. Objective The purpose of this scoping review was to provide an overview of existing literature on the psychosocial roles, risks, and boundaries of AI chatbot use, identify the concept of reliance and its connections to attachment, dependency, and parasocial interaction, and highlight gaps in the current literatures, particularly regarding potentially maladaptive reliance on AI chatbots. Methods This scoping review followed the Arksey and O’Malley framework with modifications by Levac et al., using the Population-Concept-Context (PCC) framework to guide the inclusion criteria and PRISMA-ScR guidelines for reporting. The review focused on young adults aged 18 to 29 years and synthesized literature on AI chatbot use and reliance-related psychosocial outcomes. Searches were conducted in ScienceDirect and Scopus. Results A total of 864 records were identified during the search, with only 3 studies meeting the eligibility criteria for inclusion in the review. Although the included studies examined various psychosocial aspects of AI chatbot use, none of them directly and consistently addressed maladaptive reliance. Conclusion With the limited findings of eligible studies examining reliance and dependence on AI chatbots among young adults, further research is necessary to identify how reliance develops, its potential psychosocial risks, and when AI chatbot use shifts from being beneficial to potentially maladaptive reliance.