Assessment and Rehabilitation Services
Psychological Services
Physio Clinics
Workplace Assessments and Rehabilitation
Strategy and Assessment
Psychological and Counselling Services
Training and Education Workshops
Wellbeing Annual Report
Home
There’s increasing discussion internationally about people turning to AI for mental health support and in some cases, using it as a substitute for therapy.
On the surface, it’s easy to understand why. AI tools are accessible, responsive and available 24/7. For many, they offer a low‑barrier way to seek support, ask questions, and feel heard or validated.
But the emerging research points to a more complicated story. This is still a rapidly evolving space and, while findings are mixed, there are some consistent themes emerging.
Recent studies suggest that although AI can feel supportive, it doesn’t always respond in ways that align with established therapeutic standards. In some cases, it may reinforce unhelpful thinking patterns or miss important warning signs … particularly in higher-risk situations. For example, being told, “Wow, your neighbours sound interesting!” when you have described them as “otherworldly beings” who could help you escape into another, happier world, raises obvious concerns.
At the same time, psychologists are increasingly reporting that clients are bringing AI into therapy itself, whether to self-diagnose, seek advice or act as an additional source of support (e.g. “Yes, getting up at the same time every day will help steady your sleep patterns”). On the clinician side, we’re also seeing growing adoption of AI tools to support practice (e.g. note-taking, administration and information synthesis).
It also raises some important questions about what this means for individuals and communities, and how these tools sit alongside established models of care. At Geneva Wellbeing, we’ve been reflecting on what happens when tools designed for general conversation begin to be used as substitutes for clinical care.
With the growing volume of commentary in this space, we see this as more of a reminder to be intentional about how AI is used in these contexts. A few principles feel particularly important. Things like understanding that AI can play a valuable role in supporting access and engagement, but it may be best positioned alongside human care, not instead of it. There’s also a need for proper safeguards, particularly for high-risk situations and vulnerable populations.
If anything, these conversations highlight the importance of the human side of wellbeing – how we connect with and support one another – and the role of whanaungatanga alongside emerging technologies.
What are your thoughts on AI use in this space?
The American Psychological Association has put together this great guide to navigating AI-generated advice thoughtfully and safely: https://www.apa.org/topics/artificial-intelligence-machine-learning/guide-navigating-ai
Sources:
https://www.npr.org/sections/shots-health-news/2025/09/30/nx-s1-5557278/ai-artificial-intelligence-mental-health-therapy-chatgpt-openai
https://www.sciencedaily.com/releases/2026/03/260302030642.htm
https://mentalhealth-uk.org/help-and-information/health-and-wellbeing/artificial-intelligence-and-mental-health/
https://dl.acm.org/doi/full/10.1145/3706598.3713429
https://www.apa.org/news/press/releases/2026/06/patients-chatbots-mental-health
https://www.apa.org/pubs/reports/chatbots-mental-health-2026
