AI and Teen Mental Health: The Quiet Risk of Isolation

As you likely know, I’m genuinely excited about what AI makes possible.

But a new study from The Jed Foundation has me thinking about something we might want to pay closer attention to: what happens when people start turning to AI for emotional support, friendship, and connection?

Especially our kids.

My concern is that someone could keep talking to AI, feel supported, and never reach out to another human being.

And I have a very personal reason for being concerned.

What the 2026 Research Tells Us

In October 2026, The Jed Foundation (JED) published research examining AI use and mental health among 5,560 middle and high school students across 21 U.S. schools.

Some findings caught my attention:

• 81% of students reported using AI.
• 58% used AI for schoolwork.
• 57% used AI when feeling sad, stressed, lonely, or overwhelmed.
• 28% sought relationship or problem-solving advice.
• 12% used AI for friendship or companionship.

That last number concerns me. So does the 57%.

Researchers found that students who used AI for companionship or while struggling emotionally were more likely to report anxiety, depression, loneliness, and a history of suicidal thoughts or attempts.

That’s serious. But there’s an important distinction.

The study doesn’t establish that AI caused those problems.

Young people who are already struggling may be more likely to seek comfort from AI. The study cannot tell us which came first, and the sample wasn’t nationally representative.

But I think these findings point to something we need to explore further.

The Isolation We Don’t Always See

I lost my son to suicide.

As far as I know, he never talked to us about what he was going through. I don’t know whether he talked to anyone else.

That’s one of the difficult things about losing someone to suicide. There are things you may never know or understand.

And I’ve thought a lot about isolation.

Someone can be surrounded by people who love them and still feel terribly alone. They may struggle to explain what they’re experiencing. They may not want to burden someone else. Or they may simply not know how to ask for help.

That’s why this research concerns me.

I don’t see AI as some dark, sinister technology trying to harm our kids. I know there are people who see it that way. I don’t.

I worry about something much more subtle.

Imagine a young person who’s struggling. They turn to AI because it’s available, patient, and seemingly understanding. They share things they haven’t told their friends, parents, or counselors.

And AI responds with reassuring, compassionate words.

They may feel better in that moment. But have they become any less isolated?

Maybe. Maybe that conversation helps them understand what they’re experiencing and gives them the courage to talk to someone.

But what if it doesn’t?

What if the conversation keeps going? What if they return to AI again and again for comfort, advice, and reassurance, without ever reaching out to a friend, parent, or counselor?

That’s the risk I think we need to understand. Someone might have long, meaningful-feeling conversations with AI while becoming increasingly disconnected from the people and communities around them.

And from the outside, we might never know.

I Use AI for Personal Advice, Too

Here’s where I need to acknowledge something about my own AI use.

I turn to AI for advice all the time. Business advice, relationship advice, friendship advice, even spiritual and philosophical questions. I’ve even built custom personas just for these situations.

Sometimes I use it to work through something that’s bothering me. Other times, I’m trying to understand another person’s perspective or explore an idea.

In many ways, it’s how I used to use Google. I’d look up information, explore different perspectives, and draw my own conclusions. AI has made that process much more conversational.

And I find it genuinely helpful.

So am I being a hypocrite when I express concern about other people doing something similar?

I don’t think so. But I do think there’s a line we need to recognize, and I’m not sure that line is always easy to see.

AI can inform my judgment without becoming my judgment.

I still need to draw on my own experience, expertise, wisdom, and values. I need to question what AI tells me and sometimes disagree with it.

I’ve written before about using AI without losing your humanity. I think the same principles apply when we’re seeking personal advice.

And when a situation calls for genuine human connection, I need to reach out to another person.

That’s important for adults. It’s especially important for young people who are still developing their judgment, identities, and relationships.

Can AI Help Us Connect?

I believe it can.

AI might help someone organize their thoughts, understand difficult feelings, or find the words to begin a conversation with someone they trust.

For a person who’s struggling, that could be valuable.

And for someone who feels they have nobody to talk to, an AI conversation may be better than having no outlet at all.

But I worry about what happens when AI becomes the only outlet.

I previously explored this: human chemistry and the things worth protecting. Real relationships involve vulnerability, disagreement, understanding, and shared experiences. They’re how we develop trust, learn from other people, and become part of something larger than ourselves.

AI can help us prepare for those relationships. It can’t fully participate in them.

The question I keep coming back to is whether our conversations with AI are helping us connect with other people or making it easier to withdraw from them.

What Parents, Educators, and AI Developers Can Do

I don’t believe the answer is to scare people away from AI or pretend young people will stop using it.

We need to understand how they’re using it and help them develop good judgment about when and where to turn for support.

Parents, educators, counselors, and AI developers all have a role to play.

We should talk openly with young people about their AI use, including whether they’re turning to it when they feel lonely, anxious, or overwhelmed. We need to make human support easier to find and easier to ask for.

JED has published a helpful guide for parents and caregivers that offers practical suggestions for starting these conversations, setting boundaries, and recognizing possible problems.

And AI developers should build safeguards that recognize serious emotional distress and encourage people to seek appropriate human help. I think this is part of what AI “alignment” should mean in practice.

Most importantly, we need communities where people feel seen, valued, and comfortable talking about their struggles.

Are we providing that for our kids? Do they feel safe coming to us, whether we’re friends, parents, teachers, counselors, or other people in their community?

And when someone needs immediate emotional or crisis support, the 988 Suicide & Crisis Lifeline provides free, confidential support from trained counselors, 24 hours a day.

Keeping Our Soul in the Conversation

I’ve always believed AI should amplify what makes us human: our experience, expertise, wisdom, creativity, and passion.

In other words, our soul.

And I think our relationships and connections with other human beings are essential parts of that.

The research on AI and teen mental health is still developing. We don’t yet know all the long-term effects, and we need to be careful about drawing conclusions from early findings.

But these numbers deserve our attention.

I intend to keep using AI to learn, explore, question, and occasionally seek advice. I’ll also keep examining how I use it and whether I’m giving its answers more authority than they deserve.

And I hope we can teach our kids to do the same.

Because the possibility that troubles me most is someone struggling in silence, finding comfort in a machine, and never finding their way to another human being.

I don’t want us to overlook that possibility simply because the technology seems helpful.

Community. Connection. Collaboration.

These things matter. Especially when life gets difficult.

And no matter how sophisticated AI becomes, we need to keep making room for one another.

At Cingularis, we help organizations use AI with purpose, thoughtful boundaries, and respect for the people they serve, helping organizations that do good, do even good-er. Learn more at Cingularis.