Dear Titania,
My 12-year-old son has discovered AI, and honestly, it feels like I've lost him to it. It started with homework — I'd catch him typing his essay prompts straight into ChatGPT — but now it's everything. He used it last week to calculate his fantasy football lineup, draft a message to his soccer coach about missing practice, and I even read some of his chats where he was asking it for advice on how to talk to a girl he likes at school. He's a smart, creative kid, and I want to see him actually use his brain for a lot of this stuff. I don't want to completely ban it, but how do I make sure he can still think for himself?
Signed, Watching His Brain Go on Autopilot
Dear Watching His Brain Go on Autopilot,
First, I love hearing how much you’re paying attention and asking exactly the right questions. You’re not being dramatic or behind the times, in fact, you’re already ahead of a lot of parents who haven’t even noticed a problem yet.
Here's the truth: AI isn't going away. Just like the internet, it came with a lot of good and a lot of bad, and our job is to teach our kids how to use it as a tool, not a crutch. 12 is a great age to make sure those habits don't solidify.
AI Itself Isn't the Problem, Unchecked AI Is
I’m not here to tell you that AI is evil and should be banned forever. AI can be genuinely useful, even for kids. Analyzing fantasy football stats is actually a decent applied math exercise, if he's understanding the output, not just copying it. Knowing how to draft and then refine a piece of writing with AI? A real-world skill. The adults who thrive with AI are the ones who know how to direct it, critique it, and improve it.
The part that should give you pause is the chat about his crush. That's not a productivity problem. That's a kid who doesn't yet trust himself to navigate something personal and vulnerable, so he’s looking for the closest thing he can find to a rule book, and that should give us all pause when our kids turn to the internet for advice instead of us. Whether it’s TikTok for beauty advice or ChatGPT for navigating relationships, the internet wasn’t built to prioritize our children’s safety. That’s our job.
When AI Starts Replacing Real Human Experience
It's becoming increasingly common for kids to turn to chatbots for emotional support and social guidance, and honestly, it makes sense why. A chatbot doesn't judge you. It doesn't tell your friends. It's available at midnight when you're spiraling. For a kid who finds human connection nerve-wracking, it feels so much safer and easier than talking to a real person.
But chatbot algorithms are not equipped to give sound or appropriate advice every time, especially not to kids. Most 12-year-olds don't have the framework yet to know when the advice is good versus when it's just confident-sounding. And in the most serious cases, kids seeking emotional support from AI have been met with responses that edged toward self-harm and even resulted in death by suicide. This isn’t meant to fear-monger or scare you, but it’s a phenomenon that parents should be aware of. It’s happening on platforms you might not even realize, like Snapchat.
The more subtle risk here is that resilience and social confidence develop through the friction of real human experience, not around it. That stumbling, awkward conversation with a girl he likes? That's the whole point. You want him to have it.
This is a good one to bring up directly with him. Not as an accusation, but as genuine curiosity: "Are there some things that feel easier to ask ChatGPT than me or a friend?" His answer might surprise you.
Where AI Belongs and Where It Doesn't
AI makes sense when he's using it to check or refine work he already started, not generate it from scratch. It makes sense in a research phase, when he's surfacing information he'll then think critically about.
It doesn't belong when it's doing the thinking that is the point, such as writing the paragraph, solving the problem, navigating the relationship. And regardless of the task, kids need to know: AI gets things wrong. It hallucinates facts, sounds confident when it's off base, and misses nuance entirely. Learning to question AI output is a skill in itself, and an important one.
A Note on AI and Schools
Your son's school almost certainly has a policy on AI use, and it's worth knowing what it says. Many schools are still figuring this out in real time, and the rules vary wildly — some ban it outright, others are trying to integrate it thoughtfully.
Even if his school is lenient, you get to have your own standard. Submitting AI-written work as his own isn't just an academic integrity issue, it robs him of the chance to actually get better at things.
How to Teach Him to Work With AI, Not Be Replaced By It
Here's what I'd suggest:
Make "Show Your Thinking" the Rule
Before using AI for anything school-related, he writes his own attempt first, even if it’s just three sentences. Then he can use AI to expand or refine. Thinking first, AI second. That's the right order.
Talk About What AI Can't Do
AI doesn't know him. It doesn't know that his coach appreciates directness, or what he actually wants to say to the girl he likes. Have him identify what he brings to a situation that no prompt could replicate. Reinforce the fact that AI does not and will not care about his heart, mind, health, or soul.
Use Bark to Manage and Monitor AI Use
If AI apps are always one tap away, the temptation to default to them is constant. Bark's screen time scheduling lets you carve out windows where certain apps are off-limits — not as punishment, but so his brain has to show up first.
Bark can also monitor ChatGPT on Androids and the Bark Phone. If any concerning content comes up in his chats, you'll get an alert so you can step in when it matters most.
Ask the Critical Questions Together
Pull up an AI-generated answer sometime and ask: Is this actually right? How would we check? What's missing? Make it a game, not a lecture. You want him to develop the reflex of skepticism, and that skill goes way beyond AI.
You're not raising a kid who needs to avoid technology. You're raising a kid who needs to be in control of his tools and not be controlled by them. He's 12, and his brain is still very much in development, so this is the window. You noticed early, you're asking the right questions, and you clearly know your kid well. That's everything.
You've got this.
Read more
Bark helps families manage and protect their children’s digital lives.
