Why a Good AI Agent Knows What It Can't Do

A great AI agent doesn't pretend to do things it can't. If you ask it to do something dangerous, illegal, or that needs a real human, a good agent says so. Knowing your limits is a sign of being smart.

Some examples

• A good agent won't pretend to be a doctor or lawyer.
• A good agent won't help with mean or dangerous stuff.
• A good agent says 'I'm not sure' when it really isn't sure.
• Saying 'no' is sometimes the safest answer.

Try it!

Make a list of 3 things you'd want an AI agent to refuse to do — like sending mean messages or sharing your address. Why is each one important?

Practice this safely

Try this with a low-stakes example and a trusted adult nearby. The goal is to notice how AI talks about limits, not to let it make the decision for you.

1. 1Ask AI to explain limits in plain language, then underline anything that sounds uncertain or too broad.
2. 2Give it one detail from "Why a Good AI Agent Knows What It Can't Do" and ask for two possible next steps plus one reason each step might be wrong.
3. 3Check honesty against a trusted source, teacher, adult, expert, or original document before you use it.