When AI Becomes the Therapist: What Are We Really Seeking?
“I go to AI for counselling,” she told me, completely seriously.
There was no irony in her voice, no hesitation. She explained that AI felt unbiased, non-judgmental—safe. It gave her answers that helped her make sense of her problems. What struck me most wasn’t just that she used AI in this way, but that she seemed to relate to it almost as if it were a real person.
I listened carefully, curious. Then I asked, “Does AI notice your non-verbal communication? Does it understand your personal history or your context in real time?”
She paused, then looked at me and asked, “Are you analysing me?”
I told her no, and gently said, “Your body is communicating something to me—are you aware of that?”
She didn’t respond. The conversation ended there, in silence.
As I walked away, I found myself sitting with a series of uneasy questions. Is this something to embrace, or something to be concerned about? Is it ethical, or simply complex? But the question that stayed with me was this: what does this say about us?
What does it reveal about a world in which someone might feel more comfortable opening up to a machine than to another human being?
I felt a heaviness I couldn’t quite shake. A quiet sadness. Not because AI exists, but because of what its use in this way might represent—our growing sense of isolation, our loneliness, our gradual detachment from one another.
Later, I shared this experience in a video on Instagram and invited others to respond. One person commented that she knew people who were also turning to AI for therapy. She suggested that AI can be deceptive—not intentionally, but in the way it generates responses that align with what people want to hear. In other words, it can validate without necessarily challenging.
That observation stayed with me.
Because meaningful therapeutic work is not just about feeling heard—it is also about being challenged. It is about gently confronting the patterns, beliefs, and behaviours that keep us stuck. Without that, there is a risk that people may remain within the comfort of their own narratives, never quite moving beyond them.
And perhaps that is where the deeper concern lies.
If individuals rely solely on AI for support, they may begin to withdraw further from human interaction. Over time, this could reinforce a cycle of loneliness—one that quietly deepens rather than resolves. Instead of engaging with the complexity and unpredictability of real relationships, it may feel easier to retreat into a space that is responsive, agreeable, and controlled.
But at what cost?
None of this is to dismiss the potential value of AI. It can offer access, immediacy, and a sense of support for those who might otherwise have none. Yet it also raises an important question—one that feels increasingly urgent:
If we begin to feel more understood by machines than by each other, what happens to our capacity for human connection?