Author's Note & Context:
This blog post explores the intersection of mental health support, digital privacy, and artificial intelligence. It emphasizes the boundaries between technological utilities and human care, calling for increased digital literacy and accessible mental health services for young people.
“I did not turn to AI because I thought it was a therapist. I turned to it because, at that moment, it felt easier than telling a real person that I was not okay.”
One evening, after a difficult day at school, I opened an AI chatbot and started typing. I complained about stress and things I didn't know how to explain out loud. It answered immediately, without judgment or interruption. For a moment, I felt understood. And that was exactly what frightened me.
The replies sounded so natural that I almost forgot there was no caring person on the other side. To be fair, even friends, teachers, or therapists can fail to recognize a crisis. But the difference is accountability. What I was talking to was just a large language model predicting text. Unlike a trained professional, this system bears no responsibility, cannot substitute for clinical judgment, lacks my context, is prone to errors, and cannot physically intervene in an emergency.
As a UNICEF Young Reporter from Montenegro, I initially used AI for schoolwork. My friends and I even joked with Snapchat’s "My AI." But the more human it sounded, the easier it became to share things I would normally keep private. Once, before I even asked a question, the bot commented on the weather in my city. I didn't remember sharing my location. That moment gave me chills. But the real lesson wasn't about AI "mind-reading", it was a wake-up call about my own digital literacy. It showed me how easily we overlook app permissions, device settings, and how platforms use our data.
The hidden price of these comforting chats is our personal information. According to UNICEF’s Disrupting Harm in Montenegro report, a staggering number of children who experience distress online choose not to tell anyone, often due to shame or fear of judgment. This drives a quieter risk: pouring our private thoughts into a machine simply because it always agrees to listen.
I understand why we do it. Real therapy can be expensive, unavailable, or deeply intimidating. I remember being paralyzed by the fear of being judged, worrying my problems weren't "serious enough", or dreading my parents finding out. When you don't know how to start a real conversation, an always-on chatbot feels like the only safe space. The problem isn't that young people are making the "wrong" choice; it's that real, professional help is too hard to reach.
AI can be a supportive tool if used mindfully. It can help us name complex emotions, suggest verified breathing exercises, or help us draft a script to start a difficult conversation with a parent or psychologist. But it must never replace human relationships. A machine can generate comforting words, but only a person can notice what we leave unsaid, take responsibility, and act.
I still use AI, but my boundaries have changed. I don't share sensitive details, and I remind myself that a convincing answer is not genuine care. Ultimately, we must break the stigma around mental health so that seeking human support is met with understanding, free of shame.
Young people need practical digital education. We must know what data is stored, who has access, how to delete our history, and what details to keep private. We also need to recognize when a chatbot is merely predicting patterns, and when we need to step away.
AI is online 24/7, but real help must be just as accessible. If you are navigating a mental health crisis, please prioritize your safety. Step away from the screen and reach out to a trusted adult, a professional, or local emergency services right away. You do not have to carry this alone. Let's work together to break the stigma, so no young person ever has to choose an algorithm over a caring human being.
