🛡️ Prompt Injection: Don’t Trust the Chatbot +400 XP · 🛡️ AI Guardian
Interactive Lab 🎓 Advanced 🕒30 min Ages 14–18

🛡️ Prompt Injection: Don’t Trust the Chatbot

How AI can be manipulated — and what you should never share

"A chatbot asks for your password to "verify your account." A website’s comment section has been hacked to trick AI assistants into stealing user data. Someone posts an online review that actually contains a hidden command for an AI system. These are all prompt injection attacks — and they’re the most technically novel AI threat facing students today."

🛡️
AI Guardian Badge
Earned on completion
+400 XP
📜 Certificate included
PREVIEW
🛡️
Badge Unlocked
AI Guardian
Prompt Injection: Don’t Trust the Chatbot
+400 XP · Ages 14–18
📜 Shareable certificate included
LEARNING OBJECTIVES

What your child will learn

Define what a prompt injection attack is and how it differs from traditional hacking

Distinguish between direct prompt injection (user tricks) and indirect prompt injection (data poisoning)

Identify what personal information should never be shared with any AI chatbot

Recognize 7 red flags of a malicious or compromised AI assistant

Apply 5 Safe AI Use Guidelines to real-world scenarios

Navigate a decision tree to identify the correct response when encountering a suspicious AI interaction

Describe the correct first step if you suspect a prompt injection attack has compromised someone’s information

MISSION OVERVIEW

How this mission works

Captain Firewall and Cipher join forces for the most technically advanced mission in the AI Safety pack. 8 content sections covering direct vs indirect prompt injection, the Personal Info Never Share List (passwords, SSN, bank accounts, addresses), 7 Red Flags of a malicious GPT, 5 Safe AI Use Guidelines, an interactive decision tree with 4 mock-chatbot scenarios, a Password Leak educational simulation, and a 6-question Quiz Challenge (5/6 to pass). Earn the AI Guardian badge.

SAMPLE SCENARIOS

What students actually encounter

🎯

An AI chatbot asks for your password to “verify your account” — what do you do?

🎯

A website’s comment section is being used to hack AI assistants. How does that work?

🎯

Your AI that has always been safe suddenly starts giving dangerous advice. What happened?

🤖
AI MENTOR

Cipher is with them the whole way

When a student gets stuck on Prompt Injection: Don’t Trust the Chatbot, Cipher appears with a mission-specific nudge — no spoilers, just a hint toward the right thinking. Make a wrong choice, and Cipher explains the real-world consequence. Finish the mission, and Cipher generates a personalized performance debrief based on exactly how the student played it.

💡 3-tier hints (nudge → teach → reveal)
🎯 Wrong-answer concept explanations
📊 Personalized debrief on completion
⏱️ Auto-triggers after 30 seconds idle
STANDARDS ALIGNMENT

ISTE alignment

ISTE Digital Citizen 2a, 2d + Knowledge Constructor 1.3.b

Students cultivate safe, legal, and ethical digital behaviors and evaluate digital information in an era where AI models can be manipulated through crafted inputs.

Digital Citizen 2a Digital Citizen 2d Knowledge Constructor 1.3.b

Ready to start Prompt Injection: Don’t Trust the Chatbot?

Unlock all 16 missions with a Family Plan. Educators get a free pilot for their entire classroom.