danielsauder

The online course How to use AI and LLMs for Offensive and Defensive Security from 2015 makes a short introduction to relevant topics regarding Artificial Intelligence (AI) and Large Language Models (LLMs) and their usage in the field of cyber security. These topics include theory about LLMs like prompt engineering or tokens. Then it goes on to explain different tools that can be used locally or with remote services such as ChatGPT. After the introduction, the course goes on to show several use cases for offensive security as well as for defensive security.

Although the course is short with only 4.5 hours of video material, it gives a solid and quick overview of the topics mentioned. For example, to myself the section about installing and using LLMs locally is extremely useful, I currently use LM Studio with several different models and will probably also try other tools mentioned in the course, as well as LLMs that could also be used for creating images and short videos.

However, there is one downside to the course; the last update was in April 2025, and for the quick development in the field of AI that is already a very long time. This means that some of the topics discussed in the course might already be outdated. But in conclusion I found the course helpful, because it is very hands-on.

In the meanwhile I setup an AI machine to run LLMs locally with eight GPUs. Next blog post to come.