970x125
I have spent nearly 30 years as a clinician, and I believe that psychotherapy, at its best, is one of the most powerful forces for human change that exists. Yet the profession is sleepwalking into a transformation it is not prepared for. Therapists are very good at helping our clients face hard truths. It is time we extended that same courage to ourselves. Here are four points that I believe deserve our immediate discussion:
1. Clients are forming healing alliances with chatbots, and the outcomes data is starting to back them up.
For decades, the therapeutic alliance has been the cornerstone of our professional identity. Not a specific modality or technique, but the relationship itself. The bond between therapist and client is the single strongest predictor of positive therapeutic outcomes across every form of treatment we have studied. It is precisely what we point to when someone asks what human therapists offer that no app ever could.
Except that people are now reporting experiencing that alliance with artificial intelligence (AI). In a 2025 survey of adults with diagnosed mental health conditions who had used an AI chatbot for mental health support, 38 percent rated it as more beneficial than traditional therapy. A 2026 randomized controlled trial (RCT) found that a therapy chatbot produced a 51 percent reduction in depressive symptoms. That same year, a second RCT of nearly 1,000 university students found that an AI platform produced greater reductions in anxiety and greater improvement in life satisfaction than face-to-face group therapy. Furthermore, this is a snapshot of a single moment in time. Large language models (LLMs) mature faster than humans do, and those available today are already superior to the versions used in these studies. What feels like a distant threat today is likely to become an immediate one faster than many anticipate.
My own experience may be relevant here. I have used AI to address some of my own personal challenges. Being a therapist, I came to it knowing how to ask the right questions. Even so, I was not expecting it to help the way it did. I also find myself feeling something unexpected: gratitude. I am accustomed to feeling grateful to the humans who have helped me along the way. Feeling it toward something that is not human is new, strange territory for me.
2. The 50-minute hour was never a clinical necessity.
There is nothing therapeutically sacred about 50 minutes. That structure emerged in the early days of psychoanalysis as a practical way for therapists to see patients sequentially and has persisted in part because insurance companies built reimbursement around it. Clients, meanwhile, are now engaging with AI that is available any time they need it, for as long as they need it, with no gap between sessions where crises can deepen or insights can fade. Of course, the weekly 50-minute hour has value in creating a space for deep work. Nonetheless, this tradition may be modified by insurance companies and chatbots alike. The hybrid therapy models coming to market are largely not built around a 50-minute hour. They offer AI support between sessions, which handles a significant portion of what clients previously needed face time for, and then shorter, more targeted human contact. Less human time, at lower cost, with AI filling the gaps. From an insurer’s perspective, that’s a definite win. From some clients’ perspective, it may be a win as well. But from a therapist’s perspective, it represents a fundamental renegotiation of what the clinical hour is worth, and will likely have downstream impact on the profession as a whole.
3. The profession is not being honest about what we actually ask of our clients.
Think for a moment about what the typical client must navigate just to access care. Finding a therapist who they like, is accepting new patients, takes their insurance, and has an opening that fits their schedule can be a frustrating and lengthy experience. Then comes the endless paperwork—HIPAA forms, therapy contracts, and now the Good Faith Estimate as mandated by the No Surprises Act. Arranging childcare or time off work. And the list goes on. It’s hard to be honest with another human being. Therapists know that clients are always, to some degree, managing how they are perceived. We know not to take clients’ reports of alcohol use at face value, for example, and we are used to clients disclosing critical information months into their treatment. And then there’s cost. Paying hundreds of dollars per session, potentially out of pocket, for a relationship that ends if they move or lose insurance. Being charged for missed appointments, even if they feel their excuse was legit and unavoidable. The impact of a therapist’s yawn. In sum, therapy can be a greater bother than we often acknowledge.
A chatbot eliminates every single one of those barriers. It never has a waitlist. It does not charge for missed appointments. It will adapt its voice, cultural frame, or therapeutic style to whatever the client prefers. It is available anytime, anywhere. None of this makes the chatbot better. But it is a genuinely significant competitive advantage that we are ignoring.
4. Competition drives prices down in every market. Ours will not be the exception.
There is no industry in which the introduction of a cheaper, scalable, always-available competitor has left the incumbent pricing structure intact. Not music. Not travel. Not journalism. Not retail. Plus, the richest companies ever to exist in human history are now pouring resources into AI mental health tools. Insurers are watching the outcomes research carefully. They are already exploring paying for “measurable outcomes” rather than simply reimbursing the hour, and hybrid models that combine minimal human oversight with AI-delivered care are increasingly attractive to them financially. The logical consequence is that therapist salaries, particularly at the entry level, will face downward pressure.
This concerns me deeply. I am not worried about established clinicians with full practices and decades of credibility. I am worried about therapists entering this field right now and those at mid-level who are not specialized. Therapists have taken on graduate school debt, completing low-paid internships, building practices from scratch, all on the assumption that the landscape they were trained for will exist throughout their career. We are not acknowledging that the economics of this profession are probably about to change in ways we cannot fully predict.
Therapeutic Alliance Essential Reads
That omission is not kindness. It is the professional equivalent of a parent in denial. We have sat with clients whose lives were shaped by what their own parent refused to see about themselves or their worlds. What the parent denies, the child inherits. The newer therapists entering this field are in that position with us right now. They are the children of a profession that is looking away. What we will not reckon with, they will be left to absorb alone. We need to extend to our younger colleagues the same honesty and tenderness we extend to our clients every day.
Of course, no one can see the future, and I am no exception. But discussing these very real possibilities can only empower us as we head into the future. In a follow-up post, I will explore what therapists can actually do. There are real options. But none of them are available to a field that has not yet looked honestly at the problem.
