LAB 372 – Defending Python AI Applications from System Prompt Leakage (NEW)

Course Overview

Learners will uncover how internal system prompts can be exposed and used to manipulate LLM behavior. The lab guides users in isolating, obfuscating, and securing prompt templates against leakage.

After completing this lab, learners will have the knowledge and skill to:

• Detect exposure of hidden system prompts
• Isolate internal instructions from user-visible content
• Randomize and obfuscate prompt structures
• Prevent user inference and prompt injection chaining