What is it about?
This article proposes a new way for psychotherapists to use artificial intelligence: not to treat patients or make clinical decisions, but as a tool for thinking about their own work after a session has ended. The framework, called SADAR, treats AI as a kind of reflective partner — a "third presence" alongside the therapist and patient — that can prompt the clinician to notice their own emotional reactions, blind spots, and assumptions. Rather than giving answers, the AI asks questions that widen the therapist's self-awareness, a process the authors call working through a structured three-step reflection: taking a position, being prompted through dialogue, and critically integrating what emerges. The approach is illustrated with two clinical examples. Throughout, human professional judgment and traditional supervision remain central: the AI is meant to support the therapist's thinking, never to replace it.
Featured Image
Photo by Vitaly Gariev on Unsplash
Why is it important?
Artificial intelligence is entering mental health care quickly, and much of the debate focuses on whether it can diagnose or treat patients directly — a prospect that raises serious clinical and ethical concerns. This article shifts the question. Instead of asking AI to do the therapist's job, it explores whether AI can help therapists do their own job better, by strengthening the self-reflection and awareness of unconscious dynamics that lie at the heart of good clinical work. This matters because reflective supervision is often scarce, and clinicians frequently work without a space to examine their reactions to difficult cases. By keeping AI firmly in a supporting role and outside direct patient care, the framework offers a safer and more realistic entry point for this technology in psychotherapy. It is a proposal meant to open discussion and invite testing, not a finished or validated method — and that is precisely its purpose: to steer the conversation about AI in therapy toward the clinician's inner work.
Perspectives
As a clinician, I have long been wary of the idea that a machine could sit in the therapist's chair. What interests me is the opposite possibility: that AI, used carefully and kept away from the patient, might help me see myself more clearly. The most demanding part of our work is not knowing what to do, but recognising what we ourselves bring into the room — our reactions, our resistances, the things we would rather not notice. SADAR grew from the wish to create a disciplined space for that kind of self-examination, without pretending the machine understands anything. I am fully aware that this is a proposal, not a proven tool, and that the risks — over-reliance, the illusion of insight, the erosion of human supervision — are real and must be studied. I offer it not as an answer but as an invitation to think together about how this technology can serve the reflective heart of psychotherapy rather than hollow it out.
PhD in Neuroscience Walter Paganin
StudioPsicologiaSignorini
Read the Original
This page is a summary of: Artificial intelligence as a digital analytic third: the SADAR framework for reflective supervision in psychotherapy, Frontiers in Psychology, May 2026, Frontiers,
DOI: 10.3389/fpsyg.2026.1690291.
You can read the full text:
Contributors
The following have contributed to this page
