The following is a guest article by Jessica Crain, Co-Founder and Chief Operating Officer at Mindful Therapy Group
A microphone in the therapy room would probably make payers happy and help clinical and billing teams sleep better at night. It would also push some clients to look for care elsewhere and some clinicians to walk away from any setting that records their sessions.
That is the reality of AI in mental health. Technology vendors promise that one more “solution” will fix access, documentation, burnout, and quality. But what I hear from our 2,500+ providers is simpler: don’t turn therapy into a permanent, machine-readable transcript just to support the system.
Digital tools undoubtedly have a real place in this work. They can help people find quality care for their needs and help practices keep their doors open in a hostile reimbursement environment. But the closer they get to the therapy relationship itself, the more the costs start to outweigh the benefits.
Consider ambient listening and transcription tools used during therapy sessions. These products are marketed to “free” the therapist from note‑taking. In the room, patients disclose painful, personal, and stigmatized truths. Knowing that every word is being recorded changes what some are willing to say. It also keeps the provider from using their own judgment about what belongs in the medical record. This poses real privacy concerns, undermines trust, and prevents full and honest disclosure. Mindfulness in therapy is a clinician’s ability to stay present with a client’s pain without flinching or reaching for a distraction. That presence strengthens the alliance. Research consistently finds that a strong alliance between therapist and client is one of the best predictors of good outcomes across therapeutic approaches, which is the ultimate goal.
The problem isn’t just privacy or data handling, though those matter. It’s what happens to trust if the client starts to feel the real audience is the tool, not the therapist. If we are solving for human connection, why does the human need a listening device to do their job? What I hear from our own providers is that they worry any use of AI during sessions will eventually train the models that could replace them. But instead of banning every tool outright, we need to design a system that centers providers and their relationships with clients.
However, rejecting every use of AI on principle would be its own kind of irresponsibility. The administrative burden on therapists, especially those taking insurance, is heavy and rising. Payers are already using sophisticated analytics and AI to scrutinize notes and claims. If we respond with manual audits and good intentions, we will lose, and providers will absorb the cost in clawbacks, delayed reimbursement, and burnout.
This is where technology should work hard: in the plumbing, not in the room. At Mindful Therapy Group, which is overwhelmingly insurance‑based, we have seen what happens when documentation doesn’t line up with payer standards. Claims are denied, payments are delayed, and access suffers. Manual review barely kept up with audits. Today, by embedding an AI‑driven chart‑review system into our workflows and EHR, we are building toward reviewing roughly 10,000 charts a month without adding staff.
This system never listens to sessions. It only sees signed notes, with identifiers restricted to what is necessary, and flags potential gaps between what is documented and what payers expect. Clinicians can ignore suggestions, amend notes, or simply adjust future documentation. The tool does not practice therapy or make clinical decisions. It helps defend against payer scrutiny, not to judge the quality of human work.
Admittedly, some of our clinicians felt blindsided, questioning whether AI had any place in their practice at all. Others worried that using AI to “battle” payer systems would move us further away from the mission of psychotherapy. We heard them. We created an opt‑out path, tightened how we scope and redact data, and clarified our position: if AI touches their notes, it will be in the back office, not in the room, and they have a say in how it shows up.
The line for us is this: AI belongs wherever it makes care easier to find, to pay for, and to sustain. That includes billing, claims, documentation quality, and business insight. It does not belong in any role that quietly turns human disclosure into data exhaust for a model, or that replaces the hard parts of therapy: judgment, challenge, and accountability.
Early research on consumer chatbots is already showing why that matters, with publicly available “therapy” tools endorsing harmful or ill‑advised suggestions in a significant share of distressed teen scenarios. Surveys also suggest that anywhere from one in eight to one in four adolescents are already turning to AI chatbots for mental health support, often without adults realizing. Most of these tools still lack the guardrails we take for granted in real care: reliable crisis detection and escalation, clear boundaries and disclosure that “this is not therapy,” transparent data use and storage, and any meaningful informed consent. That is a much bigger conversation, and one I suspect we will all be having as more families discover that the first “listener” their teenager turns to is not a human at all. As a parent of teens myself, I do not just think about that as an operator. I think about it sitting at the dinner table, wondering if and when I should talk to my own kids about who they are really talking to about their struggles. Whether it is a chatbot on a teenager’s phone or an audit tool in a practice like ours, the same question applies: Does this technology serve human relationships, or does it start to replace them?
Mental healthcare does not need more technology for its own sake. It needs discipline about where technology genuinely serves the work and where it distorts it. The future platforms that earn clinicians’ and clients’ trust will be the ones running on a thick layer of AI in the back office and a very thin one at the point of care. The human relationship has to stay at the center. Everything else is optional.
Jessica Crain is the Co-Founder and Chief Operating Officer at Mindful Therapy Group, where she leads operational strategy and multi-state growth across one of the largest provider-centered behavioral health platforms in the region. Over the past fifteen years, she has built the infrastructure, processes, and teams required to scale care in a fragmented and highly regulated environment, supporting thousands of clinicians across multiple markets. With a background as a Registered Nurse at the University of Washington Medical Center, Jessica brings a grounded understanding of clinical care to her operational leadership, approaching system design through both a systems and human lens. She is particularly focused on how technology, disciplined execution, and provider-aligned models can reshape access to mental healthcare in the U.S.
No comments:
Post a Comment