For the therapists and clients who want to understand our approach
If you have landed here, you probably already know how hard it has become to get mental health care. We want to start there because it is the reason Therapy Ally exists. The country is in the middle of a mental health access crisis, and it is not improving. Tens of millions of adults experience a mental illness each year, and close to half of them receive no treatment at all. About half the country lives in an area with a shortage of providers, and the need for care is growing far faster than the number of clinicians who can meet it. Behind those numbers are real obstacles: the cost of a single session, the months spent on a waitlist, the absence of any provider within reach, and the quiet belief that you are supposed to handle it on your own.
What troubles us most is that the systems meant to solve this, government, healthcare, and the technology industry, have not built anything that gives those millions of people meaningful access to care. There are apps, pilots, and studies, but not yet a real path to help for the person who has none. That gap is the problem we set out to address, and it shaped our choice about how to build.
We decided to focus on strengthening the relationship between a therapist and a client, rather than replacing it with a conversation between a person and a machine. We made that choice for three reasons:
1. It is the only approach currently substantiated by evidence.
2. It is the approach the American Psychological Association stands behind.
3. It is the one that fits the laws now taking shape across the country.
The rest of this page explains what that means in practice and why we think it matters for you.
Therapy Ally is an AI support tool for the time between sessions. Your therapist sets it up and directs it, and it supports the work you do together during the hours and days when you are not in the room. It is not a chatbot standing in for your therapist, and it is not a place to receive a diagnosis or treatment. It is a way to carry the momentum of a session forward, to practice what you and your therapist are working on, and to stay connected to that work between appointments.
If you are a therapist, that means you stay in control. You decide how the Ally you set up supports each client, within the approach and boundaries you set, and you remain responsible for the care. If you are someone whose therapist uses Therapy Ally, it means the support you receive from us is an extension of your therapist’s work, not a substitute for it. The relationship at the center of your care stays human.
There is a tempting idea that therapy is mostly a set of techniques, and that if you could package those techniques well enough, software could deliver them. Decades of research point the other way. The most studied ingredient in successful therapy is not a particular technique; it is the alliance, the trusting, collaborative bond between a client and a therapist working toward shared goals. Across hundreds of studies and tens of thousands of clients, that bond is one of the strongest predictors of whether therapy helps. And when researchers look closely, a therapist’s skill at building that bond matters more than almost anything else, which tells us the relationship is not a generic backdrop; it is a craft that lives in a trained human being. We built Therapy Ally to support that craft, not to imitate it.
Therapists have long described something that passes between two people in a room, and it turns out to be measurable. When a client and a therapist are truly attuned, their bodies begin to move in concert, heart rhythms, breathing, and the subtle signals of the nervous system start to synchronize. Researchers have found that this physical synchrony tracks with how much empathy a client feels and with the strength of the alliance. It is part of how a person who is overwhelmed can borrow the steadiness of someone calmer until they find their own footing.
An AI cannot take part in that exchange. It has no heartbeat to fall into rhythm with yours, no breath, no nervous system, no body in the room or across the video link. However warm its words may sound, it is not doing what two human beings do together. That is not a flaw we expect a better version to fix; it is simply the nature of the tool.
So we use the tool for what it is good at, supporting the work, and we leave the work itself where it belongs, between you and your therapist.
You have probably seen apps that describe themselves as AI therapists or that offer AI therapy. We understand the appeal. They are available at any hour, they do not judge, and they cost little or nothing. If you cannot reach a human at all, we are not here to shame you for turning to one. But we think you deserve to know what they are and what they are not. An app that markets itself with the word therapist while quietly noting, somewhere in its fine print, that it is not a replacement for a licensed clinician is asking you to trust a word it will not stand behind. We would rather be straight with you. Therapy Ally is not your therapist. It is support for the relationship with the person who is.
This is also the direction the law is moving. Several states, including Utah, Nevada, Illinois, and most recently Vermont, have passed rules on the use of AI in mental health, and they all draw the same line. They make room for AI that assists a licensed professional, and they close the door on AI that tries to act as a licensed professional. We built Therapy Ally on the right side of that line because a licensed clinician is always the one directing the care and remaining responsible for it.
We do not want to pretend that we have all the answers. For the millions of people who cannot reach a therapist at all, the real choice is not between an app and a human; it is between an app and nothing, and the evidence on whether a well-built tool beats nothing in that situation is genuinely mixed. Some careful studies show a real benefit. Other tools, especially the general-purpose ones, have caused real harm. We take both seriously, and it is part of why we built the way we did. Therapy Ally is made for the situation where you do have a clinician, and our whole purpose is to make that relationship stronger and more continuous, not to become a lonelier substitute for it.
We will always be clear about what Therapy Ally is. We will not call our software a therapist, and we will not ask you to trust it with something it was never built to hold. If you are a clinician, we would be glad to show you how it works inside your practice. If you are here because your own therapist uses it, we hope this helps you understand why and what it is doing on your behalf. The relationship is the point. Everything we build is in service to that.
Baldwin, S. A., Wampold, B. E., & Imel, Z. E. (2007). Untangling the alliance-outcome correlation: Exploring the relative importance of therapist and patient variability in the alliance. Journal of Consulting and Clinical Psychology.
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.
Flückiger, C., Del Re, A. C., Wampold, B. E., & Horvath, A. O. (2018). The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy.
Heinz, M. V., Jacobson, N. C., et al. (2025). Randomized trial of a generative AI chatbot (Therabot) for mental health treatment. NEJM AI. Dartmouth College.
Horvath, A. O., Del Re, A. C., Flückiger, C., & Symonds, D. (2011). Alliance in individual psychotherapy. Psychotherapy.
Kleinbub, J. R. (2017). State of the art of interpersonal physiology in psychotherapy: A systematic review. Frontiers in Psychology, 8, 2053.
Koole, S. L., & Tschacher, W. (2016). Synchrony in psychotherapy: A review and an integrative framework for the therapeutic alliance. Frontiers in Psychology, 7, 862.
Marci, C. D., Ham, J., Moran, E., & Orr, S. P. (2007). Physiologic correlates of perceived therapist empathy and social-emotional process during psychotherapy. Journal of Nervous and Mental Disease.
Norcross, J. C., & Lambert, M. J. (2018). Psychotherapy relationships that work III (APA Division 29 task force). Psychotherapy.
Tschacher, W., & Meier, D. (2020). Physiological synchrony in psychotherapy sessions. Psychotherapy Research.
Wampold, B. E., & Imel, Z. E. (2015). The Great Psychotherapy Debate: The Evidence for What Makes Psychotherapy Work (2nd ed.). Routledge.
Woebot postpartum depression and anxiety randomized controlled trial (2025), reported in 2 Minute Medicine.
U.S. mental health treatment and workforce data: SAMHSA national survey estimates (2024); HRSA behavioral health workforce projections through 2033; Milliman / NPR reporting on shortage areas and out-of-pocket psychotherapy costs (2021-2023).
Utah HB 452 (2025), effective May 7, 2025. Nevada AB 406 (2025), effective July 1, 2025. Illinois HB 1806, Wellness and Oversight for Psychological Resources Act, signed August 4, 2025. Vermont Act 156 (H.816), signed June 17, 2026.
Reported chatbot safety failures: research presented at the 2025 ACM Conference on Fairness, Accountability and Transparency, and reporting via CNN (2025) and MindSite News (2025) on dangerous responses from general-purpose AI chatbots.