A Child's AI Use Is a Traffic Light

Originally written in Traditional Chinese: 孩子用 AI，不是開關，是紅綠燈

Texas is asking a court a yes-or-no question: should minors be allowed to use ChatGPT? The state wants the answer to be no.

On October 23, the Texas Attorney General sued OpenAI and its CEO Sam Altman in a 67-page petition. The state alleges the products were released without sufficient parental controls, age verification, or safety testing. Among the relief it seeks: barring minors from ChatGPT, and requiring a court-appointed independent evaluator to approve new models before they launch in Texas. OpenAI said it has worked with experts to mitigate risks and build additional safeguards, and that "Governments have an important role to play in setting robust safety standards for AI."

Two clarifications. These are requests, not a ruling; the court has decided nothing. And Texas is not the first: on October 14, eight other states each filed their own suits against OpenAI.

I am not weighing in on who is right. What interests me is the shape of the question. "Can kids use it?" is an on/off switch. A court has to ask it that way, because it governs millions of children. A family does not. A family can ask a finer question.

A child's use of AI is not a switch. It is a traffic light.

The same child, the same tool: looking things up, asking AI to quiz you, checking an essay you already wrote. Versus pasting in a whole assignment, using it during a test, passing it off as your own. Those are different acts. A blanket ban removes the good uses along with the bad. Full freedom hides the bad ones from view.

In my course I give parents an AI traffic light. Green: extend research, ask AI to guide you with questions, review finished work. Yellow, ask an adult first: completing part of an assignment, translating a whole passage. Red: handing over the whole assignment, using it in a test, pretending you wrote it. When the color is unclear, one question decides it: "If AI does this, will your brain get stronger or weaker?"

To be clear, a traffic light does not let AI companies off the hook. Product safety should not rest on parents alone. But a court's answer may take a long time, and your child will open an AI tool tonight.

Try this: sit down with your child, list three things they used AI for this week, and decide the color of each together. Disagreement is fine; the argument is where judgment gets practiced. Educators can ask students to color-code their own AI use. Professionals can do the same with their work: which tasks make you stronger when AI helps, and which ones are you slowly forgetting how to do?

The court will eventually answer "can kids use AI?" The more durable answer is a child who has learned to ask, "Will this make me stronger or weaker?"

Get the Human-First practice card

Four steps, ten minutes, with copy-ready AI prompts. The card and my weekly letter are currently written in Traditional Chinese.

A Child's AI Use Is a Traffic Light

Get the Human-First practice card

Follow in English

References