Benefits of CCBT Chatbots: A Contemporary Review

Digital Health · Andreas Grivas, MSc Clinical Psychologist · July 22, 2023

Benefits of CCBT Chatbots: A Contemporary Review

Mental health chatbots, also called conversational agents, are programmes that talk with users through text or voice and guide them through psychological exercises. Many are built on the principles of Cognitive Behavioral Therapy (CBT): they help people check in on their mood, notice automatic thoughts, spot thinking traps such as catastrophizing, plan rewarding activities and practise problem-solving. Some follow carefully scripted, rule-based dialogues written by clinicians; newer ones use artificial intelligence (AI) to generate more flexible responses. Part of the motivation for chatbots came from a well-known weakness of earlier web-based CBT programmes: they work, but many people stop using them. A conversational format was proposed as a more engaging, always-available way to deliver the same techniques (Fitzpatrick et al., 2017).

One of the first randomized trials tested Woebot, a fully automated agent delivering CBT-based self-help. Seventy young adults aged 18 to 28 with self-reported symptoms of anxiety and depression were randomly assigned to two weeks of conversations with Woebot (up to 20 sessions) or to an information-only e-book on depression. Participants engaged with the chatbot about 12 times on average over the two weeks. The Woebot group showed a significant reduction in depressive symptoms, while the information group did not; anxiety decreased in both groups among those who completed the study (Fitzpatrick et al., 2017). The study was small, short and unblinded, but it showed that CBT can be delivered conversationally in a way that young people find acceptable.

Real-world data have added to this picture. In an evaluation of Wysa, an AI-based conversational app, anonymous users who reported symptoms of depression were divided according to how much they used the app. Those who used it more showed a greater average improvement in depression scores than lower users (5.84 versus 3.52 points on the PHQ-9, effect size 0.63), and 67.7% of user feedback described the experience as helpful and encouraging (Inkster et al., 2018). Because this was an observational study without a control group, it cannot show that the app caused the improvement, and the authors themselves called for validation in larger samples over longer periods.

Meta-analyses now allow a broader view. An early systematic review of 12 studies found weak evidence that chatbots improved depression, distress, stress and acrophobia, no significant effect on subjective well-being, and conflicting results for anxiety; only two studies assessed safety at all (Abd-Alrazaq et al., 2020). A later meta-analysis of 11 randomized trials in adults with depressive or anxiety symptoms found that chatbot-delivered psychotherapy significantly improved depressive symptoms (g = 0.54) (Lim et al., 2022). The largest review to date, covering 32 randomized trials and 6,089 participants, found small but significant short-term effects on depressive symptoms (g = 0.29), generalized anxiety (g = 0.29), general distress (g = 0.33), stress (g = 0.24) and quality of life or well-being (g = 0.27). However, for most outcomes the long-term effects were not statistically significant. Personalisation and empathic responses were identified as key ingredients, and longer interaction with the agent was linked to larger effects (He et al., 2023).

Recent analyses point in the same direction. A meta-analysis of 18 randomized trials with 3,477 participants found improvements in depression (g = 0.26) and anxiety (g = 0.19) that were most evident around eight weeks of use, but no substantial effects at three-month follow-up (Zhong et al., 2024). Another review compared rule-based and large language model chatbots: rule-based chatbots had a small but significant effect on depression (g = 0.27), whereas evidence for chatbots based on large language models was not yet statistically conclusive, with very wide confidence intervals because few trials were available (Du et al., 2025). Taken together, the benefits of CBT chatbots appear real but mostly small to moderate and short-lived without continued practice.

So what are the genuine benefits? Chatbots are available at any hour, including late at night when worries tend to peak. They offer a low-threshold, private first step for people who are hesitant to talk to someone because of stigma, cost or waiting lists, barriers that have been widely described in this field (Inkster et al., 2018). They can deliver structured CBT techniques consistently, remind users to practise between sessions, and help people track patterns in their mood. For someone already in therapy, a chatbot can reinforce skills learned in sessions. Digital tools like these can complement, but not replace, professional care.

The limitations matter just as much. Many trials are small, short and compared chatbots with waiting lists or information leaflets rather than with active treatments, and the risk of bias in early studies was often high (Abd-Alrazaq et al., 2020). Effects tend to fade after treatment ends (He et al., 2023; Zhong et al., 2024). Safety has been studied far less than effectiveness, and chatbots may misunderstand users or miss signs of risk. Privacy is an important consideration, since conversations contain highly sensitive information. Chatbots are not appropriate as the main source of help for people with severe depression, suicidal thoughts or self-harm, psychosis, bipolar disorder, eating disorders, substance dependence or recent trauma.

It is time to seek professional help if your symptoms persist or worsen despite several weeks of self-help, if they interfere with your work, studies or relationships, or if you notice hopelessness or thoughts that life is not worth living. A clinical psychologist or psychiatrist can provide a thorough assessment and an individualised treatment plan. If you are in crisis or thinking about harming yourself, contact your local emergency number immediately; a chatbot is not a crisis service.

In summary, CBT-based chatbots are a promising and accessible way to learn and practise evidence-based coping skills, and research consistently shows short-term reductions in depressive and anxiety symptoms. The size and durability of these effects remain modest, and safety and long-term outcomes need more study. Used as a first step or a complement to care, rather than a substitute for it, chatbots can play a useful role in modern mental health support.

References

  • Abd-Alrazaq, A. A., Rababeh, A., Alajlani, M., Bewick, B. M., & Househ, M. (2020). Effectiveness and safety of using chatbots to improve mental health: Systematic review and meta-analysis. Journal of Medical Internet Research, 22(7), e16021. https://doi.org/10.2196/16021
  • Du, Q., Ren, Y., Meng, Z. L., He, H., & Meng, S. (2025). The efficacy of rule-based versus large language model-based chatbots in alleviating symptoms of depression and anxiety: Systematic review and meta-analysis. Journal of Medical Internet Research, 27, e78186. https://doi.org/10.2196/78186
  • Fitzpatrick, K. K., Darcy, A., & Vierhile, M. (2017). Delivering cognitive behavior therapy to young adults with symptoms of depression and anxiety using a fully automated conversational agent (Woebot): A randomized controlled trial. JMIR Mental Health, 4(2), e19. https://doi.org/10.2196/mental.7785
  • He, Y., Yang, L., Qian, C., Li, T., Su, Z., Zhang, Q., & Hou, X. (2023). Conversational agent interventions for mental health problems: Systematic review and meta-analysis of randomized controlled trials. Journal of Medical Internet Research, 25, e43862. https://doi.org/10.2196/43862
  • Inkster, B., Sarda, S., & Subramanian, V. (2018). An empathy-driven, conversational artificial intelligence agent (Wysa) for digital mental well-being: Real-world data evaluation mixed-methods study. JMIR mHealth and uHealth, 6(11), e12106. https://doi.org/10.2196/12106
  • Lim, S. M., Shiau, C. W. C., Cheng, L. J., & Lau, Y. (2022). Chatbot-delivered psychotherapy for adults with depressive and anxiety symptoms: A systematic review and meta-regression. Behavior Therapy, 53(2), 334–347. https://doi.org/10.1016/j.beth.2021.09.007
  • Zhong, W., Luo, J., & Zhang, H. (2024). The therapeutic effectiveness of artificial intelligence-based chatbots in alleviation of depressive and anxiety symptoms in short-course treatments: A systematic review and meta-analysis. Journal of Affective Disorders, 356, 459–469. https://doi.org/10.1016/j.jad.2024.04.057