The Night I Realised AI Felt Emotionally Safer Than People

A few months ago, I found myself sitting in bed, typing a message into ChatGPT with tears running down my face.

It wasn’t a research question.

It wasn’t a to-do list.

It wasn’t a carefully worded email.

It was a confession..:something I had been carrying alone. And what scared me most wasn’t the confession itself. It was how natural it felt to send it to a machine rather than to a person.

I told myself it was practical. I didn’t want to worry my mother. I didn’t want to burden my friends. I didn’t want to risk opening up only to be met with silence or misunderstanding. ChatGPT offered me something that felt like emotional safety: a predictable response. A steady tone, a kind of controlled comfort. But as I listened to a podcast episode about someone who had formed an unhealthy emotional bond with AI — a young boy who felt more seen by a machine than by anyone in his life — something inside me stopped.

I asked myself: how did we get here? How did we get to a world where loneliness feels easier to manage with an algorithm than with another human being? Our grandparents had neighbours who checked in. Our parents had friendships woven into their daily lives. And we — the digital generation — have a dozen platforms and almost nowhere to put our fears.

I think part of what’s happening is that vulnerability has started to feel like a performance. We curate it, we ration it, we share it in “soft” versions that still protect us from rejection. But a machine requires no performance. There is no relationship to negotiate. No fear of being “too much.” And that can feel deeply comforting…until it isn’t.

The truth is: AI can soothe but it cannot hold. It can respond but it cannot repair. It can comfort but it cannot care. Healing still needs human beings.

As I prepare for season two of Mindful Mondays — where we will continue delving into digital intimacy, AI reliance, and emotional safety — I keep asking myself: what would our lives look like if vulnerability felt safe again? What would happen if we rebuilt our communities in a way that allowed for softness, honesty, and imperfection? I don’t have the answers yet, but I know the question matters. And maybe, instead of judging ourselves for turning to machines, we should use it as a the massive none sign that it so clearly is: we miss each other, we need each other, and we are not meant to do life alone.

If AI feels safer right now, let that be information, not shame. This is the beginning of a bigger conversation…

…and I’m grateful you’re here.