A practitioner's grounding in red teaming an autonomous artificial intelligence (AI) agent, distinct from the model reliability testing covered in evaluating-models-for-security-tasks. You will learn why granting an agent tools, memory and a goal changes the object under test, how to scope an exercise around the agent's real granted tools and permissions, the three recurring failure modes worth testing for, goal misgeneralization, tool misuse and persistence or resistance to being stopped, how to safely contain a red team exercise in an isolated environment, the documentation standard a finding needs when the fix is a permission or architectural change rather than a code patch, and an honest look at why this remains a young, evolving discipline without a single settled methodology.
Nothing. Every course, exam, and certificate on the catalog is free — including retakes. All you need is a free Safeguard account.
None. The flagship course, Safeguard Certified Practitioner, is a beginner-level course — basic familiarity with how software is built helps, but every exam question is answerable from the lessons themselves.
You can retake it after a 24-hour cooldown, as many times as you need. Retakes are free, and each attempt draws a fresh random set of questions.
The Safeguard Certified Practitioner credential is valid for 24 months from issue. The expiry date is printed on the certificate and shown live on its public verification page. Renewing means passing the current exam again.
