The global rise of substance-related and behavioral addictions highlights the need for scalable, accessible, and evidence-based psychological interventions. Traditional forms of counseling and rehabilitation often fail to ensure timely support, continuity of care, and sufficient therapist availability. This study aims to analyze and systematize international research on the integration of artificial intelligence chatbots into psychological support for addiction treatment.
Methods: Literature was searched in PubMed, Scopus, Web of Science, and Google Scholar (2017-2025) using predefined keywords; inclusion focused on chatbot interventions for substance-related or behavioral addictions reporting empirical outcomes or validated therapeutic frameworks. Representative models were analyzed: Woebot/W-SUDs, Tess, Chatbot-Assisted Therapy (CAT) for methamphetamine use, Quin for smoking cessation, the simulation-based LLM system ChatThero, and a field-defining systematic review.
The analysis demonstrated that chatbot-based interventions are feasible and acceptable for individuals with substance-use and behavioral addictions. Reported outcomes include reduced craving intensity, improved emotional regulation, and increased engagement in treatment. CAT showed measurable clinical improvement through toxicology data and retention rates, while Quin and ChatThero demonstrated enhanced adaptation to motivational interviewing and relapse-prevention frameworks.
Analysis revealed that AI chatbots demonstrate high feasibility and acceptability among users with various types of addictions. Many studies have noted reduced craving intensity, improved emotional self-regulation, increased engagement in therapy, and enhanced motivation for change. Certain models have demonstrated clinically significant effects, including improved toxicology scores and increased retention in treatment programs.
The findings indicate that AI-mediated conversational systems can effectively complement traditional therapy by providing continuous support and promoting self-management. Further research is needed to verify long-term clinical outcomes, ensure ethical oversight, and adapt interventions to diverse cultural, linguistic contexts.
This research was supported by the Committee of Science of the Ministry of Science and Higher Education of the Republic of Kazakhstan (Grant No. BR24992927, "Integrative Study of Gambling Addiction in Kazakhstan and Multidisciplinary Strategies for Its Minimization).