LAB 371 – Defending Python AI Applications from Excessive Agency (NEW)

Course Overview

This lab explores how giving LLMs too much autonomy, such as executing privileged actions, can lead to operational and compliance risks. Learners will assess dangerous automation paths and implement strict governance.

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

• Recognize over-permissive LLM behaviors
• Restrict autonomous capabilities and scope of action
• Enforce human-in-the-loop checks for sensitive tasks
• Define operational boundaries for AI components