The demonstration shows Security GPT analyzing a polymorphic trojan that used multiple attack vectors including spear phishing, CVE exploitation (CVE-2021-34527), pass-the-hash lateral movement, rootkit installation, and command-and-control callbacks. The AI is trained on data from over 20,000 real-world devices to understand complex, multi-stage attacks across various threat categories.
