The effectiveness of AI-based cognitive behavioral therapy chatbots for anxiety and depression symptoms: a systematic review and meta-analysis

Background AI chatbots are increasingly being used to provide accessible psychological support. However, the effectiveness of chatbot-delivered cognitive behavioral therapy for anxiety and depressive symptoms remains uncertain. This study aimed to evaluate the effects of artificial intelligence chatbot-based cognitive behavioral therapy on anxiety and depressive symptoms. Method A systematic search of PubMed, Embase, Web of Science, PsycINFO, Cochrane Library, CNKI, VIP, Wanfang, and CBM databases was conducted from their inception to July 8th, 2025. Updated search on August 25, 2026. Meta-analysis was conducted using Stata version 18.0 and RevMan version 5.4. The ES for the treatment effect was calculated using the SMD . Results Ten studies involving a total of 1,301 participants. The results of the meta-analysis showed that the CBT of AI chatbots could significantly improve depressive symptoms at the end of the intervention ( d = −0.40, 95% CI : −0.62 ~ −0.17, P = 0.0005). In contrast, the intervention did not significantly improve anxiety symptoms ( d = −0.029, 95% CI : −0.145 ~ 0.087, P = 0.623), with no substantial heterogeneity detected ( I 2 = 0%). The subgroup analysis for depressive symptoms indicated significant effects were observed in participant age stratification, intervention Duration, and Control group intervention. However, some subgroup estimates were based on a small number of studies, so the data should be interpreted with caution. Conclusion This meta-analysis emphasizes that AI-based chatbots for CBT may have great potential in contributing to the alleviation of depressive symptoms. Given the small number of included studies and the high heterogeneity, the results should be interpreted with caution. More high-quality RCTs are needed in the future to further validate these findings and explore their long-term benefits. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420251089835 .