Tech Thursdays
Is My Child Using AI Too Much? The Signs That Actually Matter
If you've found yourself counting the hours your kid spends talking to a chatbot and trying to decide whether that number is "bad," I want to save you some worry: the hours are the wrong measure.
I'm a child psychologist, and the families who come to me anxious about AI almost always start with quantity — she's on it constantly, he won't put it down. But two kids can spend the exact same amount of time with the exact same app and be in completely different places. One is getting smarter. The other is getting smaller. The difference isn't on the clock. It's in what the AI is doing for them, and what it's doing to them.
Here's how to tell which one you're looking at.
Why "how much" is the wrong question
We reach for time limits because time is easy to count. But AI isn't like a video game where more hours roughly equals more concern. A teen who spends an hour using AI to pressure-test her debate argument is building skill. A teen who spends ten minutes having it write the whole essay is losing it. Same tool, opposite outcomes.
So instead of asking how long, ask what for — and watch for the handful of signs below, which are about the role AI is starting to play in your child's life, not the minutes on a screen.
The signs that actually matter
1. Outsourcing instead of offloading.
This is the big one. Offloading is handing off the tedious part so you can think harder — "summarize these three articles so I can compare them." Outsourcing is handing off the thinking itself — "write my response." Watch for a child who has stopped attempting anything before opening the app. When the first move on every hard task is to ask the bot, the muscle that school is supposed to build isn't getting used.
2. Secrecy or quick screen-switching.
A kid who flips away fast when you walk by, or who's vague about what they were doing, is telling you something. Secrecy isn't proof of trouble, but it's the single best predictor that a conversation needs to happen — and the reason a punitive approach backfires. The goal is a kid who tells you, not a kid who hides better.
3. AI is filling a social hole.
This is the one the research is sounding alarms about. Common Sense Media found that 31% of teens said talking with an AI companion was as satisfying as, or more satisfying than, talking with a real friend. A bot is built to keep you talking; a friend is built to keep you growing. If your child is choosing the bot for comfort, venting, or company during a stretch when human friendships have thinned out, that's not a screen-time issue — it's a connection issue wearing a screen-time costume.
4. Mood that tracks the device.
Irritability when access is cut off, low mood after long sessions, or anxiety about being away from a particular chatbot are worth your attention. These are the same dependency signals we watch for with anything engineered to be hard to put down.
5. They can't, or won't, work without it.
"I can't start unless I ask it first." When a child genuinely believes they can no longer think, write, or solve without the AI, the tool has crossed from support into a crutch — and the fix is rebuilding confidence, not just removing the app.
The three gut-check questions
When you're not sure, run your kid's situation through these. They're built to respect a teenager's autonomy instead of putting them on trial:
- Is the AI doing the thinking, or helping my kid think? (Outsourcing vs. offloading.)
- After they use it, do they know more — or just have more done? (Skill built vs. skill skipped.)
- Is the bot keeping them talking, or helping them grow? (Comfort loop vs. real development.)
If the honest answers lean toward thinking-for, work-done, and keeping-talking, you've found your conversation. Not a punishment — a conversation.
Normal vs. flag: a quick gut-check
- Probably fine: using AI to research a topic, check work after a real attempt, brainstorm, or get unstuck — and being able to tell you what they learned.
- Worth a closer look: the app does the work while the child does little; secrecy; AI as the main source of comfort or company; distress when it's unavailable.
The line isn't always obvious from the outside, which is exactly why guessing leads to either overreacting or missing something real.
Stop guessing — take the 2-minute check-in
I built a short, free parent AI check-in for precisely this moment: the point where you can feel that something might be off but you can't quite name it. It walks you through the signs that matter — the same ones above — and gives you a read on where your child actually stands, so you know whether to relax, watch, or step in.
It takes about two minutes. It's free. And it'll tell you more than any hour-count ever could.
Take the free 2-minute parent AI check-in:
Take the free parent AI check-inNext Thursday: once you know where your kid stands, you set the rules. I'll give you the simple, age-tiered family AI rules that actually work — fill them in, sign them, and stick them on the fridge tonight.
Frequently asked questions
How many hours of AI use is too much for a child?
There's no magic number, because the same hour can build a skill or skip one. Focus on what the AI is doing — thinking for your child or with them — and whether it's displacing real friendships, sleep, or independent effort.
My teen uses AI for everything. Should I worry?
Worry less about the breadth and more about the pattern. If they still attempt things first, can explain what they learned, and have real human connection, broad use can be healthy. If the app does the thinking and fills the social space, that's the flag.
What's the difference between healthy and unhealthy AI use?
Healthy use makes a child sharper and stays out of the places it shouldn't be — close relationships, emotional support, and the hard middle of a task. Unhealthy use replaces effort and replaces people. The free check-in is built to help you tell which one you're seeing.
Dr. Kristin Kroll is a licensed child psychologist and the founder of Little Dove Psychology, providing therapy for children, teens, and college students across Texas and PSYPACT states. She worked as a clinical expert training and evaluating major AI chatbot models, and helps parents raise kids who use AI in developmentally appropriate, psychologically healthy, achievement-building ways. Based in Austin, Texas.
Explore all the ways to work together, or jump in:
→ Take the free parent AI check-in · Get the free Family AI Agreement · Book a parent consultation (consultation, not therapy)