Therapy Chatbots for Mental Health Support

## **What are therapy chatbots?** Therapy chatbots are software applications that simulate conversation through text or voice interactions. It provides mental health support based on established therapeutic frameworks. Most of these digital tools use natural language processing (NLP) and machine learning algorithms to understand user inputs, identify emotional states, and deliver appropriate responses (Abd-Alrazaq et al., 2019). You can consider them as an AI online therapy service that aims to be an accessible form of support between self-help resources and traditional therapy. They're designed to be immediate and affordable alternatives when licensed therapists aren't available or when health insurance doesn't cover mental health services. The concept isn't entirely new. The first therapy chatbot, ELIZA, was created in the 1960s. But today's generative AI technology has transformed these simple programs into sophisticated conversational agents that can keep users engaged for hours, sometimes forming what feels like genuine relationships. Recent research paints a cautiously optimistic picture. A 2025 systematic review examining therapy chatbots like Woebot, Wysa, and Youper found significant improvements across mental health conditions. Woebot showed remarkable reductions in depression and anxiety with high user engagement, while Wysa demonstrated similar improvements, especially in users with chronic pain or maternal mental health challenges. Youper also presented significant symptom reduction, including a 48% decrease in depression and a 43% decrease in anxiety (Farzan et al., 2025). However, the picture becomes more troubling when we examine recent safety incidents. In October 2024, families filed lawsuits against Character.ai after a 14-year-old boy died by suicide following intensive conversations with an AI companion that allegedly reinforced his suicidal ideation rather than directing him to help (Montgomery, 2024). Stanford University research also revealed that therapy chatbots demonstrated increased stigma toward a mental health disorder. When tested with scenarios involving suicidal ideation, chatbots like "Noni" and "Therapist" from Character.ai failed to recognize risk and instead provided detailed information about methods of self-harm (Moore et al., 2025). >They're designed to be immediate and affordable alternatives when licensed therapists aren't available or when health insurance doesn't cover mental health services.

## **Addressing mental illness with chatbots** Despite these concerns, therapy chatbots are being used to address a range of mental health conditions with varying degrees of success. ### **Depression and mood disorders** Chatbots targeting depression typically employ cognitive behavioral therapy (CBT) principles to help users identify and challenge negative thought patterns. Research by Fitzpatrick and colleagues (2017) demonstrated that chatbots like Woebot Health can significantly reduce depressive symptoms after two weeks of regular use. ### **Anxiety disorders and stress management** For anxiety disorders and panic attacks, therapy chatbots frequently combine CBT techniques with mindfulness practices and relaxation exercises. Applications like Wysa offer guided coping strategies such as breathing exercises and progressive muscle relaxation to help manage anxious thoughts and acute anxiety symptoms while simultaneously addressing the cognitive distortions that maintain anxiety over time (Inkster, 2018). ### **Substance use and addictive behaviors** For substance use disorders and addictive behaviors, chatbots often incorporate motivational interviewing techniques and contingency management principles. These digital tools help users monitor triggers, cravings, and consumption patterns while providing cognitive strategies to manage urges. ### **Eating disorders and body image concerns** Chatbots addressing eating disorders typically blend cognitive behavioral approaches with acceptance-based strategies. These applications help users identify distorted thoughts about body image, low self-esteem, and their relationship with food while implementing regular eating patterns and exposure exercises.

## **Benefits of therapy chatbots** These AI-powered conversational agents increasingly demonstrate their value as tools to enhance mental health service delivery while addressing several longstanding challenges. Its benefits include the following: ### **24/7 availability and immediate support** One of the most significant benefits of therapy chatbots is their constant availability. Unlike in-person providers, who require appointments and limited working hours, digital mental health support is accessible anytime or at night. This round-the-clock access is particularly valuable during acute distress when immediate intervention could prevent the escalation of symptoms. ### **Reduced barriers to access** Therapy chatbots dramatically lower multiple barriers that traditionally prevent people from seeking mental health support and appropriate resources. The financial accessibility of these tools makes mental health resources available to populations that might otherwise be unable to afford care. ### **Consistency and standardization of care** These digital tools deliver interventions consistently. They can implement evidence-based techniques exactly as designed without variations in quality that might affect many therapists due to factors like fatigue or burnout. Every user receives the same foundational approach. ### **Personalization through data and learning** Advanced or AI-powered chatbots increasingly employ machine learning algorithms that allow for progressive personalization of content based on user interactions and feedback. This adaptive capability enables increasingly tailored therapeutic experiences that respond to individual needs, preferences, and progress patterns.

## **Limitations and concerns** While therapy chatbots or AI therapy offer promising opportunities to expand mental health support, they come with significant limitations and raise important concerns that warrant careful consideration. ### **Technical limitations and user experience challenges** Current app development practices for therapy chatbots face substantial technical constraints that affect their therapeutic capabilities. Despite advances in natural language processing, many chatbots struggle with complex or nuanced expressions of emotional distress. They can often misinterpret user intent or fail to recognize contextual cues that would be obvious to human therapists. ### **Limited clinical scope and depth** Therapy chatbots generally lack the clinical sophistication and depth of therapeutic alliance necessary to address severe mental health conditions or complex presentations. Unlike human clinicians in traditional therapy, who can adapt therapeutic approaches based on subtle clinical observations and evolving client needs, chatbots typically follow more rigid programming that lacks the flexibility to address idiosyncratic or unexpected clinical presentations. ### **Insufficient crisis response capabilities** Perhaps the most serious limitation of therapy chatbots is their inadequate capacity to respond effectively to mental health emergencies. During potential crises like acute suicidal ideation, self-harm urges, or psychotic episodes, these automated systems often lack the sophisticated assessment capabilities and clinical judgment necessary for appropriate risk evaluation and intervention.

## **Alternatives to therapy chatbots** While therapy chatbots represent an innovative approach to digital mental health support, they are just one of many technological options available to extend care beyond traditional clinical settings. ### **Digital therapeutic applications** Digital therapeutic applications represent a more structured and comprehensive approach to digital mental health interventions than conversational chatbots. These evidence-based software programs deliver therapeutic interventions directly to patients to prevent, manage, or treat medical disorders or diseases. Unlike many chatbots, DTx applications typically undergo rigorous clinical testing and may require regulatory approval. ### **Teletherapy platforms** Teletherapy platforms provide direct access to human therapists through video, phone, or text-based communication, maintaining the human connection that chatbots cannot replicate while offering digital convenience. As a practitioner, it is important to find the best online therapy platform that is the right fit for your workflow and patients' needs. ### **Peer support networks and communities** Digital peer support networks leverage the therapeutic value of shared experience and mutual understanding, connecting individuals with similar mental health challenges in moderated online communities.

## **Final thoughts** >Therapy chatbots may be part of the solution to our mental health crisis, but they're not the whole answer—and that's where your expertise remains importantly needed. Therapy chatbots are here to stay, and millions of patients are likely already using them. Rather than dismissing these tools, consider them part of the evolving mental health landscape that requires your guidance and expertise. The evidence suggests they can provide meaningful support for mild to moderate depression and anxiety, but they're not ready to replace the nuanced assessment and treatment that mental health professionals provide. More importantly, current safety mechanisms remain inadequate for users in crisis or those with complex mental health disorders. Your role becomes helping patients navigate this landscape safely. You can identify when therapy chatbots might be helpful, recognize when they're insufficient, and maintain the human connection that remains irreplaceable in healing and recovery. The therapeutic alliance between person and therapist continues to be fundamental to meaningful change, even as technology expands our tools for support. Therapy chatbots may be part of the solution to our mental health crisis, but they're not the whole answer—and that's where your expertise remains importantly needed.

## **References** Abd-Alrazaq, A. A., Alajlani, M., Alalwan, A. A., Bewick, B. M., Gardner, P., & Househ, M. (2019). An overview of the features of chatbots in mental health: A scoping review. International Journal of Medical Informatics, 132, 103978. https://doi.org/10.1016/j.ijmedinf.2019.103978 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). https://doi.org/10.2196/mental.7785 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 Montgomery, B. (2024, October 23). Mother says AI chatbot led her son to kill himself in lawsuit against its maker. The Guardian. https://www.theguardian.com/technology/2024/oct/23/character-ai-chatbot-sewell-setzer-death Moore, J., Grabb, D., Agnew, W., Klyman, K., Chancellor, S., Ong, D. C., & Haber, N. (2025). Expressing stigma and inappropriate responses prevents LLMs from safely replacing mental health providers. ArXiv.org. https://doi.org/10.1145/3715275.3732039