Artificial Intelligence is no longer something waiting in the wings. It’s already here, and AI in counselling and psychotherapy in ways that would have seemed unimaginable only a few years ago.
AI Is Already Changing the Therapy Profession
As therapists, we’re faced with some profound questions. What actually makes therapy effective? What is the value of sitting with another human being when a chatbot can offer instant responses, twenty-four hours a day? And perhaps most importantly, does AI threaten the future of our profession, or does it simply challenge us to better understand what makes therapy uniquely human?
The Growing Role of AI in Counselling and Psychotherapy
Many practitioners are already experimenting with AI in counselling and psychotherapy practice (I definitely am! I’ve become close friends with Claude, ChattyG and Gemini, so far). Some use it to streamline administration, manage schedules, or organise notes. Others use generative AI to explore ethical dilemmas between supervision sessions, deepen their understanding of therapeutic models or case conceptualisation, or generate ideas for client resources. There are also AI-powered mental health tools that provide psychoeducation, mood tracking, self-help exercises, and conversational support between sessions. In many ways, these developments offer genuine benefits. They can improve accessibility, reduce administrative burden, and provide support to people who might otherwise struggle to access help.
Ethical Challenges Therapists Cannot Ignore
Yet alongside these opportunities come significant risks. One of the most important concerns is confidentiality. The information we hold about clients is among the most sensitive data imaginable. Whenever client information is entered into third-party AI systems, questions immediately arise about privacy, consent, data storage, and security. Therapists have an ethical responsibility to understand exactly how these systems work and whether they adequately protect client information. There is also the question of clinical judgement. AI can identify patterns, generate summaries, and offer observations. However, therapy is not simply a pattern-recognition exercise. Human experience is nuanced, contextual, and often messy. A grieving client may appear depressed. A withdrawn client may be processing trauma. An algorithm can highlight possibilities, but it cannot truly understand the lived experience of the person sitting in front of us. AI should complement professional judgement, not replace it.
The Benefits of AI in Mental Health Care
At the same time, I don’t believe we should dismiss AI outright. Used thoughtfully, it offers real advantages. It can free therapists from repetitive administrative tasks, provide useful tools between sessions, support training and supervision, and even create innovative treatment opportunities through technologies such as virtual reality. The challenge is not whether AI has value. It clearly does.
Why Human Therapists Still Matter
The deeper question is whether AI can replicate the therapeutic relationship itself. Decades of research consistently point to the therapeutic alliance as one of the strongest predictors of successful outcomes. While techniques and interventions matter, it is the quality of the relationship that most often creates meaningful change. And relationships are fundamentally human. A chatbot cannot sit quietly with someone in profound grief. It cannot notice the exhaustion in a client’s posture, the subtle changes in their presentation, or the courage it took simply to show up. It cannot genuinely celebrate a breakthrough, share in a moment of laughter, or navigate the inevitable tensions and repairs that occur within human relationships. Most importantly, it cannot offer authentic human presence. As AI becomes increasingly sophisticated, we may eventually see systems capable of analysing facial expressions, interpreting vocal tone, and engaging in highly convincing therapeutic conversations. Some people may even prefer these systems because they are available around the clock and feel non-judgemental. But therapy has never been solely about receiving well-crafted responses. It’s my firm belief that much of its healing power comes from risking vulnerability with another person. There is something profoundly transformative about being accepted, understood, and valued by another human being who is not programmed to respond positively. The possibility of rejection exists, and yet acceptance is offered. That experience can be deeply corrective and healing. I suspect this is where AI will continue to encounter its greatest limitation.
What the Future of Therapy May Look Like
The future of counselling and psychotherapy will undoubtedly include AI. The technology is advancing too quickly and offers too many benefits to ignore. Our task is not to choose between AI and humanity, but to find an ethical and thoughtful balance between them. As therapists, our role is evolving, not disappearing. The heart of therapy remains the same as it has always been: a relationship built on empathy, compassion, trust, and genuine human connection. Whatever technological changes lie ahead, I believe that will remain the foundation upon which meaningful therapy is built.
Here are five reflective questions that invite you to think beyond the practicalities of AI and into the deeper philosophical, ethical, and relational implications for practice:
- What aspects of my therapeutic work could be enhanced by AI, and what aspects must remain distinctly human if therapy is to retain its transformative potential?
- How do I ensure that convenience, efficiency, and accessibility do not come at the expense of confidentiality, clinical judgement, or the quality of the therapeutic relationship?
- If a client begins to prefer interactions with AI over interactions with people, what might that tell us about their needs, fears, attachment patterns, or experiences of human relationships?
- As AI becomes increasingly capable of mimicking empathy and therapeutic dialogue, how would I explain the unique value of human therapy to a prospective client?
- What ethical responsibilities do I have to understand the technologies I use, particularly when those technologies have access to sensitive client information, influence clinical thinking, or shape therapeutic decision-making?
