Limits & Safety
Hallucinations, permissions, secrets, and the checks that keep AI-assisted work safe.
Models can produce plausible but false information. Important claims and code need verification.
Secrets should stay out of prompts, committed files, and logs. Use environment variables or a secret manager.
Permission checks give you control before the agent performs risky or unexpected actions.
Review and tests provide evidence that the change is correct and stays within scope.
Irreversible and high-impact actions need explicit review and approval.
Your score — Limits & Safety
/ (%)
Learn this topic through practical guides and real coding workflows inside AI Workshop.
This is a chapter of the free AI Fundamentals quiz. Answer each question, get immediate feedback with an explanation, and move on. Your progress is saved in your browser — nothing leaves your machine, and there is no signup.
We teach every topic in this quiz inside AI Workshop. Join AI Workshop to learn how to apply it in real projects.