Far more organizations use AI than build it, and that shift alone puts AI systems inside the attack surface every security professional is already responsible for. This self-paced foundational course assumes you know cybersecurity and assumes nothing whatsoever about AI or machine learning. It starts by pinning down what these systems actually are when you look at them as a defender, then covers the AI threat landscape through MITRE ATLAS, what tends to go wrong with LLM-based applications through the OWASP Top 10 for LLM Applications, and how much further the attack surface stretches once autonomous AI agents are in play.
From there everything is connected back to the NIST AI Risk Management Framework, so the individual risks resolve into a structure you can govern against. The course finishes on the practical side — the defenses and working habits that let a team absorb AI into normal security operations: knowing what AI you have, filtering what goes in and what comes out, keeping tool access to least privilege, monitoring behavior, and placing human review where a decision needs a person in it.
What You Will Learn
- Explain what makes AI/ML systems a security attack surface
- Tell traditional ML, generative AI, LLMs, and AI agents apart
- Find your way around the AI threat landscape using MITRE ATLAS
- Threat-model LLM applications against the OWASP Top 10 for LLM Applications
- Weigh the risks that autonomous AI agents and excessive agency introduce
- Put the NIST AI Risk Management Framework to work (Govern, Map, Measure, Manage)
- Apply baseline AI defenses: inventory, input and output filtering, least-privilege tool access, monitoring, and human review gates
This course is available for 365 days. You can choose when to start your access period. Once you activate, you will have 365 days to complete it (access the course materials, quizzes, projects and videos). You may request one extension of seven (7) days. Other extension requests will be evaluated on a case-by-case basis. Videos are not downloadable.
