Agentic Red Teaming: From Zero to Hero

About this course

Agentic Red Teaming: From Zero to Hero

Learn how to identify, test, and mitigate security risks in LLM and Agentic AI systems using the DeepTeam framework.

This course is designed to give you a practical understanding of LLM Red Teaming and Agentic AI Security through clear explanations and hands-on coding examples. You will learn how attackers can manipulate AI systems, how vulnerabilities are discovered, and how to build stronger defenses around LLM-powered applications.

What you will learn

Understand the fundamentals of LLM Red Teaming

Work with the DeepTeam framework for automated AI security testing

Perform Prompt Injection, Jailbreaking, Roleplay, Encoding, and other adversarial attacks

Understand single-turn and multi-turn attacks

Test for vulnerabilities such as PII Leakage, Prompt Leakage, Bias, Toxicity, Misinformation, and unsafe content

Explore Agentic AI vulnerabilities, including Excessive Agency, Goal Hijacking, Tool Misuse, Agent Identity Abuse, and Inter-Agent Communication risks

Create and test custom vulnerabilities for your own applications

Understand security frameworks such as OWASP, NIST AI RMF, MITRE ATLAS, and the EU AI Act

Implement Guardrails and mitigation strategies

Perform hands-on red teaming using Python and DeepTeam

Explore AI code scanning and automated security testing workflows

Throughout the course, you will work with practical examples involving LLMs, RAG applications, and tool-enabled AI agents.

Read the full description on Udemy on Udemy (opens in a new tab)

Prices, discounts and availability are set by Udemy. We may earn a commission when you purchase through links on this site.