An evolving part of my psychotherapy practice is hearing the dialogues and observing the use of AI in mental health practices and how shows up in the psychotherapy room. I’m on board with a low barrier access to emotional support that tends to be non-judgemental. I also think it can be an incredibly helpful tool in practicing coping strategies between sessions.
However, what feels tricky are the risks around privacy, since chatbots are not bound to the same confidentiality obligations as Registered Psychotherapists. Moreover, AI can create or contribute to a potential confirmation bias that can be tricky for clients and unintentionally exacerbate problems. It’s fundamentally biased due to the subjective information that humans provide. Chatbots tend to be overly affirming and can miss out on the broader context of someone’s history or presentation. For example, things like body language or what’s not being said.
While I’m not outright anti-AI for the folks who do find it to be beneficial, I do think that reflecting on the function the AI might be serving in our lives or its impact to our emotional wellbeing and therapeutic goals is certainly a worthy endeavour.
