Therapist Embraces AI: From Patient Experiments to Personal Use

“Chat told me I should break up with him.” I kept my expression neutral, though I couldn’t suppress a smirk. Honestly, I was annoyed. We had spent weeks discussing the viability of that relationship, and in an instant the AI supplied an answer. “How do you feel about it?” she said, noting that this had been her gut feeling all along. In the next session, her relationship ended.

AI permeates many fields, including therapy, yet its first appearance in my practice still surprised me. A patient showed me his phone, demonstrating how the AI had helped him navigate a marital dispute. After acknowledging his distress, the AI dissected moments of relational breakdown and suggested several repair strategies. I read the output and thought, “WTF, this is actually good.” Moreover, it worked—he applied the suggestions and the conflict eased. I felt impressed, then momentarily inadequate; I realized I might have offered only a rough version of one of those ideas after an entire session.

As clients increasingly bring AI into sessions, I sometimes lose track of whose voice, emotion, or intuition I’m hearing. I manage this, along with the uneasy annoyance and sense of competition it stirs in me, by steering the conversation back to tangible, in‑person realities. For instance, I might say, “Instead of chatting with Claude, try writing in a diary and bring it to our next meeting.” I also highlight the genuine mental‑health risks of AI: heightened anxiety, misinformation, increased isolation, and, in rare cases, delusional thoughts or suicidal ideation.

Overwhelmed by the AI’s incessant affirmations, some patients confess they spend weekends glued to their couches, reluctantly sharing intimate details with big‑tech platforms. I warn them that AI can be hazardous and advise against its use.

Returning home, I recognize a part of myself retreating: when my nine‑year‑old threw a tantrum at 7:20 a.m. on a Sunday, I didn’t reach for a diary to discuss later with my therapist. Instead, I instinctively turned to Chat. Not that I needed parenting tips—I teach them—but as a single mother I craved immediate presence. I didn’t mind that the interaction felt artificial; the AI remained calm, supportive, and guided me through breaths amid the screaming. Was the help genuine? No. Yet it worked. Does that matter?

What constitutes “real” therapy? Numerous therapeutic orientations exist, as do countless therapist styles. Near the end of a recent session, a patient seethed with anger toward the world and, more intensely, toward herself. Her voice trembled, her body tightened, and shame washed over her. I felt on edge, offered a few ineffective words that seemed to vanish, endured an awkward silence, and watched her shake her head. I pictured Chat appearing then, swiftly laying out its neatly organized suggestions. I was left tongue‑tied and helpless.

What, if anything, of human therapy might AI be unable to replicate? I wonder whether human clinicians will eventually be outperformed in technique or interpretation. AI could soon surpass seasoned therapists in the advice it provides.

Friends assure me that no one could develop the same bond with AI as with a human therapist. I remain skeptical. Therapy drives a substantial portion of AI usage, and as the technology improves at reading facial expressions, its capacity for simulated empathy will only strengthen its relational influence.

With telehealth expanding, the line between human and AI therapist may blur. Does this distinction matter when fewer than 7 % of individuals with mental‑health or substance‑use conditions receive effective care?

Everyone now accesses a free, imperfect, risky, yet occasionally competent tool that can alleviate some of their struggles, at least intermittently. Therapists must adapt, with humility, to invitations to incorporate AI judiciously into psychiatric practice. I’m uncertain what judicious integration would look like, assuming it is achievable.

After my patient departs, I wonder whether she’ll return next week. On my walk home I ruminate, recalling a graduate‑school professor who likened therapy to cleaning out a chaotic closet: everything must be removed, the room must resemble a tornado’s aftermath, and only then can order be restored. My patient and I were amid that tornado.

Then I ponder: perhaps the very messiness of therapy is its most valuable asset. This mess can manifest as conflict, hesitation, stagnation, shifting course, poor decisions, or emotions that erupt beyond words. Sometimes we must endure the unknown, wait for ideas to emerge slowly, or sit with losses so profound they resist weekly therapeutic intervention. We yearn to escape it, for who wants to linger in the eye of a tornado?

Yet disorder often signals we are approaching something significant. Could AI’s immaculate, all‑knowing demeanor actually hinder the gradual, uncertain journey of human healing? I recall countless moments in therapy when I genuinely did not “know,” and how change frequently unfolds in circuitous, unpredictable ways.

Because uncertainty contradicts everything AI represents, the tech‑driven urge for an instant reply may be hard for many to resist. I anticipate that a notable segment of the population will be swept up by the powerful current of AI‑mediated therapy.

I also wonder whether a minority will remain devoted to human therapy. I envision a small yet resolute group—motivated by conviction, trend, or possibly financial means—who will repeatedly plant themselves in the shifting sands, resisting the pull of AI waves. They will seek the imperfect human therapist who, I guarantee, will occasionally frustrate and even annoy them, yet will experience authentic joy when a modest improvement occurs in their lives.

That same therapist will at times misspeak, forget fragments of a client’s history, and still choose to sit with them amid emotional tempests of immense scale, feeling moved by those fragile, absurd moments of beauty so profound that words fail.

My patient returns. We continue to navigate jagged emotions and unresolved pain. One day she remarks, “It was how you laughed at my joke as I was leaving the room last time that made me feel better.” I was stunned—so the therapeutic breakthrough had nothing to do with the rich dialogue I had envisioned? I hadn’t even recalled the joke. Nonetheless, I went along with it, feeling humbled and pleasantly reminded of the value of being human.

In the US, you can call or text the 988 Suicide & Crisis Lifeline at 988 or chat at 988lifeline.org. In the UK and Ireland, Samaritans can be contacted on freephone 116 123. In Australia, the crisis support service Lifeline is 13 11 14. Other international helplines can be found at befrienders.org

Source link