This course is designed for penetration testers, red teamers, security researchers, AI security practitioners, and cybersecurity professionals who want to understand how modern AI and LLM-enabled systems can be manipulated and exploited. The course will introduce you to the rapidly evolving field of AI security by exploring the most critical vulnerabilities affecting LLM applications, including prompt injection, guardrail bypass, sensitive data exposure, and Retrieval-Augmented Generation (RAG) exploitation. You will also be introduced to industry frameworks such as the OWASP Top 10 for LLMs and MITRE ATLAS to understand how AI threats are categorized, modeled, and assessed in real-world environments.
Throughout the course, you will learn how direct and indirect prompt injection attacks work, how AI guardrails and filters are implemented and bypassed, and how attackers manipulate AI systems through instruction override, obfuscation, data extraction, and poisoned retrieval pipelines. By the end of the course, you will be able to identify, test, and exploit prompt injection vulnerabilities, assess the security posture of AI-enabled systems, and understand how modern AI workflows and integrations expand the attack surface of enterprise applications.
