Almost everyone I know is ethically against AI. It’s understandable why. We’re building technology that is changing the way we learn, create, and relate to one another. AI is changing millions of lives, yet most of us have little to no say in how these systems are designed, what values they reflect, or how they’re integrated into our daily routines. Furthermore, there are valid concerns about misinformation, privacy, environmental impact, and what it means to automate something as deeply human as conversation. So when I first heard that young people were using AI for emotional support, I was skeptical. My mind immediately went to the various online platforms and games that have been linked to tragic outcomes over the last decade.

Then I caught myself because I’ve done it too. Not because I think AI is a replacement for the people in my life. Not because I believe a chatbot understands me. But because there are moments when emotions hit so suddenly that you’re not looking for someone to solve the problem. You’re looking for someone to sit with you in it. When I’m overwhelmed, the last thing I want to hear is someone telling me I’m overreacting or explaining what I should do next. Personally, I want someone to acknowledge that what I’m feeling is real and that it’s okay to feel hurt, anxious, disappointed, or scared before trying to fix any of it.

I’ve noticed that’s exactly what AI does. It validates first.

Reading Hopelab’s report, I realized I wasn’t alone in that experience. Many young people described turning to AI because they didn’t want to burden friends, feared being judged, wanted the freedom to be honest, or needed someone who would listen immediately and without interruption. As I reflected on why I’ve done the same, I realized it wasn’t because I thought AI cared more than the people in my life. It was because there are moments when I don’t want to wonder how someone will see me after I say something messy. I don’t want one vulnerable conversation to permanently shape how someone thinks about me. I want the freedom to express exactly what’s on my mind without worrying that my words will be remembered or used against me later.

That feeling is something I think many of us relate to, and it says as much about our relationships as it does about the technology itself.

The report also raises an important concern: AI can become too validating. It can tell us what we want to hear instead of what we need to hear. In my experience, that’s been true. But I also think it’s worth asking why validation from a chatbot can feel easier to receive than validation from another person. Somewhere along the way, many of us learned that helping someone means fixing them. We rush to offer solutions, explain why they shouldn’t feel a certain way, or try to make the problem disappear as quickly as possible. It makes us feel productive. Useful, even.

Sometimes that’s exactly what’s needed. But sometimes people don’t need solutions first. They need someone willing to stay with them in the discomfort. I think that’s what AI is accidentally getting right. It cannot truly empathize, but it has learned to pause before it problem-solves, and I think that’s something we can learn to do better.

Paradoxically, the biggest thing this report changed for me wasn’t my opinion of AI. It was my perspective on human relationships.

I don’t think AI is becoming more human. I just think it’s exposing the parts of human connection that many of us are struggling with.

The irony is that AI can feel emotionally safe precisely because it isn’t in a real relationship. It won’t carry your words into tomorrow. It won’t get tired of listening. It won’t misunderstand your intentions or remember your worst moments the next time it sees you. Humans do all of those things. And yet, those same messy, imperfect qualities are also what make real relationships meaningful. I believe what makes us human is being known by each other, not just validated.

AI can create the feeling of being understood, but it can’t share in our lives, celebrate our joys, or grow with us. That’s why I don’t think the most important question is whether AI is good or bad. Responsible AI absolutely matters. We need tools that are transparent, protect privacy, and know when to encourage human support. But after reading this report, I came away thinking that the bigger question isn’t just what AI is doing. It’s what we’re missing from our relationships that makes a chatbot feel like the easier place to be vulnerable.

Maybe the real question isn’t why AI has become so good at comforting us. Maybe it’s what it would take for us to become better at comforting each other.