Learning Together as Therapy Enters the Age of AI
Artificial intelligence is entering the therapeutic space whether the profession has fully decided how to accommodate it or not.
Clients may already be using AI to understand their emotions, explore relationships, seek reassurance, interpret psychological concepts or continue thinking about material that emerged during therapy. Therapists themselves are beginning to encounter AI in documentation, education, formulation, reflective practice and clinical training.
AART-345 begins from the position that these developments deserve neither uncritical enthusiasm nor automatic rejection.
They deserve professional curiosity, observation and dialogue.
What Are We Seeing in Our Clinical Work?
AART-345 invites therapists, counsellors, psychologists and supervisors to share what they are beginning to encounter in practice.
What are clients asking AI between sessions?
What kinds of responses are they bringing back into therapy?
When does AI appear to facilitate reflection—and when might it reinforce an existing narrative?
What happens when a client experiences an AI response as unusually validating, insightful or emotionally significant?
Could AI interaction become part of transference, attachment, avoidance or reassurance-seeking?
And how might the therapist understand something that has occurred between the client and an AI system without either dismissing it or granting the technology greater psychological authority than it deserves?
These are emerging clinical questions for which the profession is still developing answers.
From Individual Experience to Shared Knowledge
AART-345 aims to create a space in which professionals can learn from one another’s developing experience.
Rather than knowledge flowing in only one direction—from technology companies towards clinicians—psychotherapists themselves have an important contribution to make.
Clinical observation can help us identify where AI-assisted reflection appears useful, where its limitations become visible, and where unexpected psychological phenomena emerge.
The objective is therefore not simply to learn about AI.
It is also to develop knowledge about what happens psychologically when human beings begin using AI within the context of their emotional lives.
Understanding the Limitations
Generative AI can produce language that appears empathic, psychologically sophisticated and highly confident while still being inaccurate, incomplete or poorly suited to the individual in front of us.
Current professional guidance consequently stresses that AI should augment rather than replace professional judgment and that clinicians remain responsible for evaluating its outputs and appropriateness.
Questions around confidentiality, data privacy, informed consent, bias, misinformation, dependency, over-validation and the potential reinforcement of maladaptive beliefs all require careful consideration.
Recognising these limitations is not an argument against innovation.
It is part of practising responsibly within innovation.
Supervision as a Place of Safety
Clinical supervision may become particularly important as AI enters psychotherapy.
Rather than leaving therapists and clients to navigate these questions independently, supervision can provide a protected professional space in which AI-assisted work is examined critically.
Supervision might explore not only what the AI said, but also:
Why did the client turn to AI at that particular moment?
What psychological need was being expressed?
How did the client experience the response?
Did the interaction increase reflection or certainty?
Did it encourage autonomy or dependency?
Did it support the therapeutic process or begin to compete with it?
What might the client’s relationship with AI tell us about attachment, transference, avoidance, reassurance or relational needs?
Are confidentiality, consent, safeguarding and professional boundaries being adequately protected?
In this sense, supervision provides something AI itself cannot provide: human professional accountability for the clinical meaning and consequences of its use.
Emerging work in professional psychology is already examining AI as an adjunct to clinical training and supervision while explicitly retaining human supervisors and professional oversight.
A Collaborative Professional Conversation
AART-345 is a developing framework.
It should therefore remain open to questioning, evaluation and revision.
The intention is to bring together clinical experience, psychological theory, research, supervision and emerging knowledge about artificial intelligence so that practitioners can learn from one another as the field develops.
Some findings may reveal genuine opportunities.
Others may reveal limitations we have not yet anticipated.
Both are valuable.
The purpose is not to prove that AI belongs in psychotherapy. It is to understand what happens when it enters the therapeutic space—and to ensure that the safety, autonomy and psychological wellbeing of the client remain central while we find out.
Observe • Share • Question • Learn
AART-345 Professional Community
Building knowledge together at the intersection of psychotherapy and artificial intelligence.