Learn the method
Read the explanations and two worked examples. Understand why an answer succeeds or fails.
Learn AI fundamentals, annotation and response evaluation. Then investigate the trust boundaries that red-team tests are designed to challenge. Clear lessons. Worked examples. Hands-on practice.

New to AI? Start with Foundations.
Already evaluating responses? Move into red-team testing.
A clear first lesson on training, inference and the human reviewer’s role.
Choose a level, or search a topic. Open a free lesson or explore a paid course.
Loading your lessons…
Real skill lives in the reason behind your answer. Every course teaches the method first, walks through two examples per course lesson, then gives you independent exercises and a final project.
Try a red-team investigation →Read the explanations and two worked examples. Understand why an answer succeeds or fails.
Review a source, an image or a conversation. Separate facts from assumptions.
Answer scored checks, write your rationale and tackle additional cases in the course lessons.
Compare with the expected outcome. Review mistakes and export your work.
Learn the fundamentals for free.
Build depth with focused red-team investigations.
The five starter lessons remain free. Buy a complete package once; no monthly subscription. Future editions may be sold separately.
Editable templates and complete course workbooks.
Use them offline, at your own pace.
Written exercises and projects use explicit self-review rubrics. They are not automatically AI-graded. Progress is saved in this browser; export work you want to keep.
These are original learning scenarios, supported by relevant public academic reading. Illustrative responses are clearly labeled. They are not recorded results from commercial models.
Objective questions use a fixed practice answer key. Written reasoning is for self-review or instructor assessment. A practice score is not a professional certification.