SEC595: Applied Data Science and AI/Machine Learning for Cybersecurity Professionals

LLMs are big these days. Chatgpt, claude, gemini, deepseek. Models seem to be appearing out of nowhere. How do we properly use these tools for maximum effect? In this talk, Jean-François Maes will go through his workflow on using LLMs to help in offensive assessments.
Many organizations have security controls in place, a defined security policy, and even conduct regular penetration testing. Yet breaches continue to occur at an alarming rate, even among those that actively engage in offensive ops. Why? The issue often lies in the gap between having security measures on paper and their actual effectiveness in the face of real-world attacks. Security controls can be misconfigured, detection mechanisms can fail, and adversaries constantly evolve their techniques.