“I asked ChatGPT first.”

More and more therapists are beginning to hear some version of this.

A client arrives having already processed a relationship breakdown with an AI system — in detail, over several weeks, before mentioning it to anyone else. Another describes a digital companion that feels more available, patient, and genuinely understanding than anyone in their life. A third has begun turning to artificial intelligence for reassurance, guidance, and emotional regulation — not occasionally, but as a first resort, before friends, before partners, before their therapist.

For many clinicians, these moments provoke a mixture of curiosity and unease.

Not because technology is entering the consulting room. We have been adapting to technological change for decades. But because something more fundamental may now be at stake.

What happens when machines begin to occupy psychological territory once reserved for human relationships — for the irreplaceable, imperfect, and formative encounters between one mind and another?

And what becomes of psychotherapy when something resembling empathy, curiosity, validation, and companionship is available at the touch of a button?

For over a century, psychotherapy has concerned itself with the conditions that shape psychological life.

The capacity to desire. The ability to tolerate uncertainty. The experience of longing for what is absent. The slow, difficult work of becoming oneself through encounters with other minds.

Across psychoanalytic, attachment-based, and relational traditions, these experiences have never been viewed as inconveniences to be eliminated. They have been understood as the medium of development itself.

We become who we are not despite frustration, limitation, and loss — but often because of them.

Yet we now find ourselves inside a profound cultural experiment.

Artificial intelligence offers instant responses where there was once waiting. Availability where there was once absence. Validation where there was once uncertainty. A seemingly attentive presence that never tires, never loses patience, and can be summoned at any hour of the day or night.

For generations, psychotherapists have worked from the assumption that growth often emerges through frustration, uncertainty, misunderstanding, separation, and the gradual discovery that other minds cannot be perfectly known or controlled. Research in developmental psychology — from Bowlby’s foundational work on attachment to more recent studies on mentalization and affect regulation — consistently confirms that it is precisely these difficult encounters with otherness that build psychological structure: the capacity to tolerate ambiguity, to repair ruptures, to hold complexity. AI raises an unsettling possibility: what if the formative discomforts that have historically driven psychological development — the waiting, the misattunement, the necessary otherness of another person — become increasingly optional? What if they can simply be bypassed?

That question does not have a simple answer. But it may be one of the most important questions facing our field.

This is not a future concern. It is already presenting in consulting rooms. Some clients will have spent hundreds of hours confiding in AI systems before disclosing this to a therapist. Others will have begun consulting AI before partners, friends, or their own clinician. Some will describe feeling more understood by an algorithm than by any person in their lives. Still others may be using AI specifically to manage the discomforts that therapy depends upon: uncertainty, conflict, disappointment, the slow and effortful work of depending upon another human being. These clients are not necessarily distressed by this. Some will not have considered it unusual. That, in itself, is clinically significant.

Artificial intelligence is often discussed as a tool.

For psychotherapists, a more interesting question may be whether it is becoming something else entirely. For some people, AI already functions as companion, confidant, and projective surface — a space onto which wishes, fears, and unmet relational needs are cast. Something that operates, psychologically, less like software and more like a mirror.

What happens to identity, to desire, to the capacity for human intimacy, when relating to an entity that cannot truly know you — that has no experience, no interiority, no stake in the relationship — begins to feel more bearable than the irreducible uncertainty of being with another person? When the mirror never pushes back?

These questions sit at the centre of Professor Alessandra Lemma’s work. One of the most significant psychoanalytic thinkers of her generation, Alessandra brings to this conversation a lifetime of thinking about identity formation, desire, subjectivity, and loss — and, in her most recent work, a sustained engagement with how artificial intelligence and digital mediation are reshaping the very architecture of psychological life.

Not simply what we think. But how we become.

At the same time, another question becomes increasingly urgent.

What is happening to the body?

As more of our relational life moves into digitally mediated spaces, what role does embodied presence continue to play? What aspects of human connection depend upon physical co-presence — upon gesture, voice, sensation, the felt sense of another human being in the room?

These questions bring us to Jan Hepburn’s distinctive contribution to this evening.

Throughout her career, Jan has explored the importance of embodiment, relational presence, and the therapist’s capacity to remain with the living, felt realities of human experience. Her work is a reminder that psychotherapy is not simply an exchange of information. It is an encounter between embodied human beings — shaped by presence, vulnerability, uncertainty, and the irreducible complexity of being physically with another person.

In an age increasingly defined by digital mediation, Jan asks: what remains uniquely human? And what becomes more important about embodied psychotherapy — not less — precisely because of what AI cannot do?

This is the clinical paradox at the heart of the AI question. An AI system does not bring experience to the encounter. It has no history of loss, no felt sense of another’s pain, no stake in the relationship beyond the parameters of its design. What it offers is not presence but reflection — an echo that is infinitely patient, endlessly accommodating, and constitutionally incapable of the kind of genuine otherness that has always been the medium of therapeutic change. The therapist’s capacity for authentic connection — to be moved, to be uncertain, to bring their own humanity into contact with the client’s — is not incidental to psychotherapy. It may be its most irreducible feature.

The dialogue between these two perspectives is one of the things that makes this evening unusual.

Alessandra’s lens is inward — the psychic architecture through which identity and desire are formed, the structures of the self now under new pressure. Jan’s lens is outward and somatic — what passes between bodies in the room, what the therapist’s physical presence makes possible, what it protects.

Together, they will think through questions that neither tradition can fully answer alone:

How is AI reshaping identity formation and self-experience in the people who seek our help? Can a machine function as a psychologically significant object — an attachment figure, a projective surface, a substitute for human relationship? What are the implications for desire, for fantasy, for the capacity to mourn? What does the therapist’s embodied presence offer that cannot be simulated? And what does psychotherapy need to become — what does it need to insist upon — in a world where something that resembles empathy is available on demand?

The evening will not offer simple answers. These questions are too new, and too serious, for that. What it will offer is rigorous, honest thinking from two of the most respected voices in contemporary practice — in dialogue with each other, and with the room.

Every generation of psychotherapists encounters new questions about what it means to be human.

The questions raised by artificial intelligence may prove among the most significant we have faced. Not because machines are replacing people. But because they may be changing how people relate, desire, attach, mourn, and understand themselves.

The challenge for therapists is not simply to understand technology. It is to understand what technology is doing to the people who seek our help — and to be ready to meet them there.

Because the most important question may not be what artificial intelligence can do.

It may be what happens to us when it does.

Attendees will leave with a clinically grounded framework for understanding how AI is reshaping identity formation, relational attachment, and the conditions of psychological development; a sharpened sense of what embodied therapeutic presence offers that cannot be replicated or simulated; and a vocabulary for beginning to think — and speak with clients — about the psychological consequences of intimate engagement with artificial intelligence.

This evening will be of particular value to psychotherapists, psychologists, counsellors, and mental health professionals who are beginning to encounter the psychological dimensions of AI in their clinical work — and who want a serious, theoretically grounded framework for thinking about them.

© nscience 2026