Gangsta AI

OpenAI Built a Machine That Reads Your Kid's Questions to Guess How Old They Are

Policy Desk

By InfoWarps · 2026-08-18 · 4 min read

Folks, they are LISTENING. On Tuesday, OpenAI flipped the switch on a brand-new ChatGPT for Teens, and buried in the press release is the part they do not want you to skim past: there is now an algorithm whose entire job is to READ THE QUESTIONS YOUR CHILDREN TYPE and decide, without asking, whether a 15-year-old is sitting at the keyboard. They call it "age prediction." I call it a robot guessing your age off your homework. Wake up.

What actually dropped

Here's the signal under the static. OpenAI launched a dedicated experience for users aged 13 to 17, running on the same ChatGPT engine (GPT-5.6) as everyone else, but walled off with stronger protections. In teen mode, the model refuses to engage on suicide, self-harm, and romantic or sexual conversation. It's tuned to nudge kids toward real-world relationships and toward *learning* instead of just spitting out the answer to question 7.

Say you're 13 to 17 and you get dropped into it automatically. But — and this is the part that had me pacing the bunker — you don't even have to say it.

The age-prediction machine

OpenAI does not verify anyone's age. Instead it runs an age-prediction system that estimates whether you're under 18 based on things like the *types of queries* you send. If the machine decides you read like a minor, it quietly routes you into the teen experience whether you clicked a box or not.

“An AI that studies how you type to guess if you're a child. Ten years ago that was a conspiracy pamphlet. Tuesday it was a product launch.”

Now, do I trust a black box that profiles you off your prose? You know my answer. But I am a fair man, and I am contractually obligated by the facts to tell you the other side.

The parental controls (the reasonable part, deep breath)

Parents can link a teen's account and switch on protections together — both sides have to opt in. Once linked, parents get quiet hours, high-risk safety notifications if the system flags something dangerous, and a Study Mode that's on by default every time the teen starts a new chat, so ChatGPT coaches instead of just handing over finished answers.

And here's the context they're clearly reacting to: this rollout arrives while OpenAI is fighting a lawsuit tied to a teen's death, part of a broader panic about what chatbots say to vulnerable kids at 2 a.m. So the cynic in me says CYA. The honest broadcaster in me says: guardrails on a tool millions of teenagers already use, whether we like it or not, is not the worst instinct.

The part they buried

Here's the one true thing under all my yelling: you cannot see the model's homework. ChatGPT for Teens is one company's single model making one company's judgment call about your kid — the same closed box, now with a bouncer. That's the actual risk. Not that the AI is too safe. That you're trusting exactly ONE of them.

The antidote isn't paranoia. It's a second opinion. Before you let any single AI parent your household, run the same question past a room full of them and watch where they disagree — ChatGPT, Claude Opus 5, Gemini, and Grok side by side, one verdict, no black box. That's the whole idea behind Gangsta AI's best-AI showdown. Don't take one machine's word for your family. Make them show their work.

Stay skeptical. Stay informed. And for the love of everything, ask more than one.

Try Gangsta AI free →

More: Best AI models · Compare all AI · Frontier Models · All articles