Knowledge BaseIntermediate

AI and LLM Security Fundamentals

In about fifteen hours you learn how AI systems built on large language models work, how they are attacked, and how to design, test, and govern them so that a fooled model can do little harm. You build a small practice assistant, watch a hidden instruction change its answer, then add guards, logging, and a staff AI use policy. Everything stays defensive.

Recommended first: Linux Fundamentals and Web Application Security, ideally also Phishing and Social Engineering Awareness.

CHF 35

Lesson 01 is free to try. Create an account to save your progress and to get the full course.

  • Certificate with a checkable ID
  • Lesson 01 is free to try
A large chat bubble drawn in thin glowing lines, with three short text lines inside it. The middle line is broken by a small bright violet gap that stands for an injected instruction. Behind the bubble, a simple neural network of three columns of small nodes linked by faint lines. A shield with a green check mark stands in front of the bubble's lower right corner. Top left, a document with a folded corner sends a dotted line into the bubble, showing untrusted content entering the model's context. Top right, a small funnel stands for an output filter.
-lessons
-modules with badges
-reading and labs
-hands-on labs
1final exam + certificate

How it works

Every lesson ends with a short checkpoint quiz. Pass it (4 of 5) and the next lesson unlocks. Finish every lesson of a module to earn its badge. Pick a pace and the timeline below gives each lesson a due date, so you always know whether you are on track.

Your pace

Your plan

What you need

One computer with Python, free tools, and an optional small local model. Every lab runs on your own machine against your own practice app, never against an AI service you do not own.

A computer with Python

A modern browser and Python 3.10 or newer on Linux, macOS, or Windows with WSL. The Ubuntu VM from Linux Fundamentals is ideal. The log lessons also use jq (sudo apt install -y jq).

A small local model, optional

Recommended: Ollama with the small model qwen2.5:0.5b (397 MB), which runs on a normal CPU without a graphics card. Without it, every lab runs in mock mode (MOCK=1), which needs only Python and gives the same result every time.

An AI security log

A text editor and a notebook, on paper or in a file. From lesson 1 on it is your AI security log: threats, decisions, test results, and policy drafts collect there and come together in the capstone.

Modules and badges

6 modules. Each badge lights up when you pass every lesson in its module, and every badge you earn is printed on your certificate.

The journey

Modules in order. Green is done, blue is where you are, amber is behind your plan, gray is still locked.

Glossary

- terms used in this course. In the lessons, underlined words show their definition when you tap or focus them.