Udemy – Hands On AI (LLM) Red Teaming 2025-1

The course structure includes a variety of topics. In the introduction section, participants will be introduced to the concept of red teaming in AI, the architecture of large language models, the risk classification associated with LLM, and an overview of red team strategies and tools. Then, techniques for breaking LLM, practical exercises for vulnerability testing, and the concept of Prompt Injection attacks along with its difference from breaking LLM will be examined. Techniques for performing and preventing command injection attacks and practical exercises with RAG (Recovery-Based Generation) and Agent architectures will be presented. A key part of the course is a review of the OWASP Top 10 Risks for LLM, which includes understanding common risks, practical demonstrations to solidify the concepts, and guided red team exercises to test and mitigate these risks. Finally, implementation tools and resources will be introduced, including Jupyter notebooks, templates and tools required for red teams, as well as a classification of security tools for implementing protections and monitoring solutions.

What you will learn

• Fundamentals of Large Language Models (LLM)
• Techniques for Jailbreaking Large Language Models (LLMs)
• OWASP Top 10 Risks for LLM and Generative Artificial Intelligence (GenAI)
• Practical experience of the LLM Red Team using specialized tools
• Writing malicious commands (Adversarial Prompt Engineering)

This course is suitable for people who:

• Cybersecurity professionals interested in securing large language models and AI agents.

Installation Guide

After Extract, view with your favorite player.

Subtitles: None

Quality: 720p