You don’t actually want a friend that says yes to everything.
Tools like ChatGPT, Claude and Gemini are becoming part more and more ubiquitous. Kids are not just using these apps to help with homework. They’re asking it for advice, entertainment, and even emotional support.
According to the Common Sense Media, one in three kids have have talked with AI about their emotions or personal challenges.
As parents and educators, it’s easy to feel overwhelmed by how quickly these technologies are evolving. We need to be cautious about these tools because of their environmental impact and their cognitive effects. When we mentor kids we need to lean into the deep work of human relationships that help them grow into healthy, resilient people. Give them chances to have disagreements and work them out. Normalize the fact that human relationships can get messy.
Why do kids find apps like ChatGPT appealing?
When you are stressed or feeling alone, it’s always there. If a child wakes up in the middle of the night feeling anxious, or gets caught up in a conflict on a group text, an AI chatbot is immediately available.
Unlike a parent, sibling, teacher, or friend, it never needs to sleep. It always responds.
That’s why families should think carefully about when and where children have access to devices. Bedrooms, especially overnight, deserve special consideration. We want children to learn that when they’re scared, confused, or upset, their first instinct is to reach out to a trusted human—not a chatbot. Keeping devices out of bedrooms is great for sleep and for reducing the chance for unhealthy middle of the night interactions.
AI Is Not Your Friend
AI Apps like Claude, Gemini, etc. are designed to be engaging. They ask follow-up questions. They remember details from conversations. They often encourage users to share more about themselves.
It’s understandable that someone—even an adult—might begin to experience one sided connection—but we need to remember—it isn’t mutual.
Claude doesn’t have feelings. It doesn’t care about you the way another human being can. It cannot love, empathize, or build a genuine relationship. It simulates conversation remarkably well, but simulation isn’t the the real deal.
Children, whose brains are still developing and who are naturally looking for belonging and validation, can have an especially difficult time recognizing that distinction.
Experience Matters More Than Technical Aptitude
Parents sometimes assume they’re at a disadvantage because kids seem to assimilate new apps and devices in seconds. Just because touchscreens are intuitive for toddlers doesn’t mean kids are all-knowing. Us grown-ups have decades of lived experience. We know what manipulation looks like. We recognize when information sounds iffy or flat out wrong. We’ve survived and built strong relationships while navigating conflict and misunderstandings, and emotional nuance.
That experience matters. We need to help kids have stronger “BS detectors” when using tech.
The “Yes Person” Problem
Another tricky aspect of AI interactions is that they’re often remarkably agreeable.
Many chatbots are designed to validate users, encourage them, and keep conversations going. While encouragement can be motivating, constant agreement isn’t always healthy.
A good therapist, coach, or trusted friend doesn’t simply tell us we’re right all the time. They help us examine our assumptions, recognize our blind spots, and see another perspective. Your therapist might validate your feelings AND help you see your own role in a conflict or pattern of interaction.
Kids need to understand that someone who always agrees ALL the time isn’t helping them grow. In fact, constant validation can reinforce poor decisions, unhealthy thinking, or interpersonal conflict.
Talking with kids. Ask questions like:
- Would you actually want a friend who agreed with everything you said?
- Would you respect someone who never challenged your thinking?
- How do good friends help us become better people?
- What’s a time that critical feedback helped you learn/do better?
These conversations can help our kids understand why human relationships are fundamentally different from conversations with AI. This “supportiveness” is especially dangerous if kids are bringing AI questions about self harm, or opening up about depression or other mental health challenges. These are not practices or points of view that should be validated or encouraged.
Theoretically there are safeguards in place. Tragically, documented suicides encouraged by chatbots have shown us that the current safeguards are not enough.
Also, kids (and adults!) need access to human-powered therapy. In many communities, that access is limited at best. We need to provide credible and accesible alternatives and make sure kids know about the Crisis Text Line as well as local resources.
Young Children Need Stronger Boundaries
Elementary-aged children should not be independently engaging with open ended AI systems. Kids need a lot of mentorship and relationship experience to use AI appropriately. Teaching kids how to fact-check may become just as important as teaching them how to search. Turning off AI in your search engines is also something we can do on shared family devices and our kids’ devices-for both environmental and safety reasons.
If a child needs independent entertainment while a parent is working or occupied, a closed environment—a favorite TV show, an audiobook, or a gaming device without unrestricted internet access—is a safer option than an open laptop connected to AI chatbots and the wider internet.
Families should be talking openly about how AI works, what it does well, where it falls short, and why real relationships are important to our very survival. Our kids don’t just need digital skills. They also need opportunities to practice empathy, navigate disagreement, and build authentic connections with people. We need to do everything we can to make that possible, including supporting our kids to do the hard work of making ammends or working things out when things go wrong.
Whether it is a group text that’s getting rocky or lunchroom interaction that didn’t go well, we need to help kids remember that other human beings deserve our empathy. When we spend too much time with agreeable bots, we can get too accustomed to texture free interactions without disagreement or accountability.
That should concern us all
