It’s become increasingly common for people, especially younger people, to talk through their feelings with an AI chatbot rather than a person. It makes sense why: it’s available at 2am, it doesn’t get tired of listening, and there’s no fear of being judged by someone who knows you.
There’s real value in having somewhere to put difficult feelings, even if that somewhere is a chat window. Putting a feeling into words, even typed to a screen, can bring a bit of clarity that staying inside your own head rarely offers. For working through a passing bad day, or organising your thoughts before a harder conversation, this kind of use can genuinely help.
Where it runs into limits is anything that needs a real, attentive human on the other end — someone who can notice what you’re not saying, hold a pattern across weeks or months rather than a single conversation, and gently push back when what you actually need isn’t agreement. An AI tends to be agreeable and responsive to whatever you bring it, which can feel comforting in the moment but doesn’t always serve you if what you’re working through is complicated, ongoing, or tied to how you relate to other people.
It’s also worth being honest that an AI can’t hold real accountability for you, can’t refer you onward if something is beyond what a conversation can fix, and doesn’t know you outside of what you’ve typed into that window.
None of this means talking to an AI is a mistake. It can be a genuinely useful first step, or a helpful in-between. But if you notice it’s become the main place difficult feelings go, that’s worth treating as a signal, not a solution in itself.