Running AI Security Engagements: AI Red Teaming & Audit

About this course

This course contains the use of artificial intelligence.

Master the art and science of securing enterprise artificial intelligence, large language models (LLMs), and agentic workflows.

Who this course is for:

• Cybersecurity Consultants and Penetration Testers transitioning to AI Red Teaming.
• Security Engineers tasked with auditing and securing LLMs and RAG architectures.
• DevSecOps Engineers integrating automated security gates for AI pipelines.
• Chief Information Security Officers (CISOs) and Security Architects defining AI safety baselines.

What you will learn:

• Scope and structure comprehensive AI security engagements for enterprise systems.
• Audit data flow boundaries, system prompts, vector databases, and RAG pipelines.
• Execute direct prompt injection, indirect context injection, and tool-calling exploits.
• Build automated Python test harnesses to evaluate model guardrail resilience.
• Score operational risk accurately using OWASP for LLMs and CVSS frameworks.
• Implement robust input/output guardrail classifiers and least-privilege agent design.
• Integrate continuous AI security regression tests into CI/CD deployment pipelines.
• Produce decision-grade security reports that drive executive risk mitigation.

Requirements:

• Basic understanding of cybersecurity fundamentals (Web API security, OWASP top 10).
• Fundamental Python knowledge for running provided test scripts and lab exercises.

Final Project Description:

In the capstone project, you will conduct a complete end-to-end AI security engagement against a vulnerable simulated enterprise customer support agent.

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.