Artificial intelligence and tobacco use: A bibliometric analysis 1997–2026

INTRODUCTION: Globally, tobacco use is a key public health issue. Exploring the impact of developing AI technologies on tobacco use and intervention will help promote the digital transformation of public health research. This study aims to map the trends, networks, and core themes of research at the intersection of artificial intelligence and tobacco use through visual analytics. METHODS: In this study, a bibliometric analysis was conducted on 335 articles on AI and tobacco use in the Web of Science core collection from 1997 to 2026. The search cutoff date was 8 April 2026. Descriptive and visual analysis were conducted on the publication trend, cooperation network, keyword co-occurrence, clustering, and burst of this literature. RESULTS: The speed of publication of literature in this field has increased significantly since 2018. In the cooperation network, national cooperation is dominated by the United States. Universities, rather than transnational institutions, significantly promote cooperation. The authors' cooperation network is more dispersed. The characteristics of interdisciplinary cooperation are obvious. The research hotspots focus on several aspects. These include machine learning prediction of smoking, dialogue AI and robot interventions, and NLP and social media monitoring. Other topics encompass the analysis of smoking-cessation behavior among youth groups, combined with AI, deep learning, and behavioral analysis; and the evaluation of tobacco characteristics, tobacco products, media, and health. Keyword hotspots and bursts reveal that the field has recently paid special attention to the intervention of cutting-edge technologies in smoking and smoking cessation behavior. CONCLUSIONS: Through data mining and visualization technology, this study reveals the overall evolution, interaction mode, and key fields of AI and tobacco use knowledge. The results provide an important framework for further research.