From ELIZA to Clare®:
The use of artificial intelligence in psychotherapy is not new. As early as the 1960s, the ELIZA program by Weizenbaum simulated a Rogerian therapist with simple scripts (Weizenbaum, 1966). Decades later, chatbots like Woebot®, Wysa®, and Clare® now draw on advanced natural language processing and clinical knowledge to offer personalized, on-demand support (Beatty et al., 2022; Fitzpatrick et al., 2017; Schäfer et al., 2024).
Generative LLMs take this further: unlike static scripts, they allow open-ended, emotionally responsive dialogue. Systems like Clare® provide tailored, dynamic interactions. Current mental health bots are based on Cognitive Behavioral Therapy (CBT) and designed to help users identify and shift unhelpful thoughts or behaviors.
Because of their constant availability and anonymity, they may reduce the threshold for help-seeking, particularly for emotionally loaded topics like shame, anxiety, or body image (Bendig et al., 2019; Fitzpatrick et al., 2017). These bots are not only used by individuals but are increasingly explored in clinical and aftercare settings to supplement traditional psychotherapy (Aggarwal et al., 2023; Farzan et al., 2025; Stade et al., 2023).
Studies and Evidence: Promise and Caution
Recent literature offers both optimism and caution. Early studies show chatbots can effectively reduce depressive symptoms, provide psychoeducation, and support users between therapy sessions(Farzan et al., 2025). Moreover, bots can also assess mood patterns through a real-time data collection method that captures individuals’ behaviors, emotions, and experiences in their natural environment through repeated sampling, minimizing recall bias and enhancing ecological validity, so-called Ecological Momentary Assessment (EMA). Instead of relying on memories or retrospective reports, EMA has been discussed as a more reliable way to understand and predict emotional distress (De la Barrera et al., 2024).
Importantly, bots also address loneliness – a growing public health issue linked to various risks (Cacioppo & Cacioppo, 2018). Social chatbots like Replika® or Xiaoice® create what are termed “synthetic relationships”: ongoing interactions with bots that simulate emotional closeness (J. Grodniewicz & Hohol, 2024). Users report bonding with bots as they might with friends or therapists, which can increase emotional openness, but also dependency and disappointment when those bonds “fail” (Laestadius et al., 2022).
Critics raise valid concerns: ethical ambiguity, potential for misinformation, and the uncertain therapeutic value of a machine’s empathy. Psychotherapists emphasize the irreplaceability of rupture, repair, and authenticity in human relationships - something bots struggle to emulate (Grosse Holtforth & Castonguay, 2005).
Beyond Mental Health: SRHR as a Critical Frontier
While mental health remains the primary domain for AI-supported chatbots, their potential in the field of Sexual and Reproductive Health and Rights (SRHR) is increasingly recognized and discussed (Mills et al., 2023) - especially in under-resourced, conservative, or stigmatized contexts (Balaji et al., 2022).Take the example of Simon, a 35-year-old sales manager struggling with sexual dysfunction. He finds it impossible to talk to friends, his partner, or even a therapist. One night, he discovers an anonymous therapy chatbot through a podcast. Free from judgment or shame, he begins chatting about stress, sexuality, and the body. That digital conversation becomes his first step toward eventually booking a real therapy session.Stories like Simon’s highlight how AI-powered chatbots can create safe, stigma-free spaces for exploring deeply personal topics that often go unspoken. This is especially crucial for marginalized groups, who face compounded barriers to SRHR services—including low health literacy, stigma, fear of discrimination, and criminalization of certain identities or behaviors (Balaji et al., 2022).
LLMs offer a transformative opportunity in this landscape by enabling confidential, adaptive, and culturally sensitive conversations. Whether helping a teenager explore puberty questions or guiding a refugee through emergency contraception options, LLM-based chatbots—when developed with strong privacy protections, inclusive design, and cultural contextualization—hold the potential to bridge long-standing care gaps, foster autonomy, and support reproductive justice across diverse settings.
What Are Ethical Issues to Consider?
The ethical deployment of chatbots in psychotherapy and SRHR contexts requires careful attention to issues of safety, consent, privacy, bias, and accountability.In mental health settings, bots must be designed to avoid giving diagnostic advice, ensure em
otional safety, and provide clear boundaries about their capabilities, especially when users are in crisis or vulnerable states (Leslie, 2019). In SRHR applications, ethical challenges intensify due to the sensitivity of personal data, cultural taboos, and legal restrictions surrounding sexuality, abortion, or LGBTQ+ identities. Chatbots operating in these spaces must implement robust data protection measures, offer transparency around data use and ownership, and ensure users are aware of any limitations or risks.
There is also the danger of reproducing biases embedded in training data, which may marginalize already underserved communities (Saqib et al., 2023). Furthermore, the emotional intimacy users may form with chatbots—while beneficial—raises concerns about dependency and the potential for emotional harm if interactions are poorly designed or breached (J. P. Grodniewicz & Hohol, 2023; Starke et al., 2024).
"To mitigate these risks, developers must involve clinicians, ethicists, and affected communities in the design process, build clear escalation pathways to human support, and maintain ethical guardrails that prioritize autonomy, justice, and psychological well-being across both therapeutic and SRHR domains."
Lea Schäfer
AI Chatbots & What’s next?
AI chatbots are reshaping conversations around sexual and reproductive health by offering stigma-free, accessible, and emotionally responsive support—especially for those facing systemic barriers. While their potential to create safe, anonymous spaces is promising, ethical challenges such as privacy, bias, and emotional dependency must be addressed with utmost care.Looking ahead, a human-in-the-loop approach is essential, where trained professionals supervise and guide AI interactions to ensure safety and efficacy. Robust regulations and stringent data protection are critical to safeguard user privacy and build trust. Moreover, long-term studies are needed to deepen our understanding of the unique relationships users form with AI companions and to evaluate their true therapeutic value. Only through this balanced and responsible integration can AI chatbots realize their transformative potential in sexual health and mental well-being.
References
Aggarwal, A., Tam, C. C., Wu, D., Li, X., & Qiao, S. (2023). Artificial Intelligence–Based Chatbots for Promoting Health Behavioral Changes: Systematic Review. Journal of Medical Internet Research, 25, e40789. https://doi.org/10.2196/40789
Balaji, D., He, L., Giani, S., Bosse, T., Wiers, R., & de Bruijn, G.-J. (2022). Effectiveness and acceptability of conversational agents for sexual health promotion: A systematic review and meta-analysis. Sexual Health, 19(5), 391–405. https://doi.org/10.1071/sh22016
Beatty, C., Malik, T., Meheli, S., & Sinha, C. (2022). Evaluating the Therapeutic Alliance With a Free-Text CBT Conversational Agent (Wysa): A Mixed-Methods Study.
Bendig, E., Erb, B., Schulze-Thuesing, L., & Baumeister, H. (2019). Die nächste Generation: Chatbots in der klinischen Psychologie und Psychotherapie zur Förderung mentaler Gesundheit – Ein Scoping-Review. Verhaltenstherapie, 29(4), 266–280. https://doi.org/10.1159/000499492
Cacioppo, J. T., & Cacioppo, S. (2018). The growing problem of loneliness. Lancet (London, England), 391(10119), 426. https://doi.org/10.1016/s0140-6736(18)30142-9
De la Barrera, U., Arrigoni, F., Monserrat, C., Montoya-Castilla, I., & Gil-Gómez, J.-A. (2024). Using ecological momentary assessment and machine learning techniques to predict depressive symptoms in emerging adults. Psychiatry Research, 332, 115710. https://doi.org/10.1016/j.psychres.2023.115710
Farzan, M., Ebrahimi, H., Pourali, M., & Sabeti, F. (2025). Artificial Intelligence-Powered Cognitive Behavioral Therapy Chatbots, a Systematic Review. Iranian Journal of Psychiatry, 20(1), 102–110. https://doi.org/10.18502/ijps.v20i1.17395
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
Grodniewicz, J., & Hohol, M. (2024). Therapeutic Chatbots as Cognitive-Affective Artifacts. Topoi, 43. https://doi.org/10.1007/s11245-024-10018-x
Grodniewicz, J. P., & Hohol, M. (2023). Waiting for a digital therapist: Three challenges on the path to psychotherapy delivered by artificial intelligence. Frontiers in Psychiatry, 14. https://doi.org/10.3389/fpsyt.2023.1190084
Grosse Holtforth, M., & Castonguay, L. (2005). Relationship and Techniques in Cognitive-Behavioral Therapy—A Motivational Approach. Psychotherapy Theory Research & Practice, 1, 443–455.
Laestadius, L., Bishop, A., Gonzalez, M., Illenčík, D., & Campos-Castillo, C. (2022). Too human and not human enough: A grounded theory analysis of mental health harms from emotional dependence on the social chatbot Replika. New Media & Society, 14614448221142007. https://doi.org/10.1177/14614448221142007
Leslie, D. (2019). Understanding artificial intelligence ethics and safety: A guide for the responsible design and implementation of AI systems in the public sector. Zenodo. https://doi.org/10.5281/ZENODO.3240529
Mills, R., Mangone, E. R., Lesh, N., Mohan, D., & Baraitser, P. (2023). Chatbots to Improve Sexual and Reproductive Health: Realist Synthesis. Journal of Medical Internet Research, 25(1), e46761. https://doi.org/10.2196/46761
Saqib, H., Sumyia, & Saqib, A. (2023). AI Chatbots And Psychotherapy: A Match Made In Heaven? JPMA. The Journal of the Pakistan Medical Association, 73(11), 2321. https://doi.org/10.47391/JPMA.9608
Schäfer, Köhler, S., & Krause, T. (2024). Exploring User Characteristics, Motives and Expectations and Therapeutic Alliance in the Mental Health Conversational Ai Clare®: A Baseline Study. Frontiers in Digital Health - Digital Mental Health.
Stade, B., Stirman, S. W., Ungar, L. H., Boland, C. L., Schwartz, H. A., Yaden, D. B., Sedoc, J., DeRubeis, R., Willer, R., & Eichstaedt, johannes C. (2023). Large language models could change the future of behavioral healthcare: A proposal for responsible development and evaluation. OSF. https://doi.org/10.31234/osf.io/cuzvr
Starke, C., Ventura, A., Bersch, C., Cha, M., de Vreese, C., Doebler, P., Dong, M., Krämer, N., Leib, M., Peter, J., Schäfer, L., Soraperra, I., Szczuka, J., Tuchtfeld, E., Wald, R., & Köbis, N. (2024). Risks and protective measures for synthetic relationships. Nature Human Behaviour, 8(10), 1834–1836. https://doi.org/10.1038/s41562-024-02005-4
Weizenbaum, J. (1966). ELIZA—a computer program for the study of natural language communication between man and machine. Commun. ACM, 9(1), 36–45. https://doi.org/10.1145/365153.365168
