RSAC 2026 video recordings are already online!

📅 Full conference program

🖥 Key talk recordings

We've selected a few talks worth checking out:

🛑AI Security in the Wild: Lessons from 2 Years of Hacking AI Infrastructure. The talk systematizes threats to AI infrastructure, focusing on data leaks caused by misconfigured cloud tokens, queue poisoning in message brokers (e.g., RabbitMQ), and arbitrary code execution during model deserialization via the insecure Python Pickle format. 🛑Chasing Ghosts: Detecting Token Abuse in the Microsoft Cloud. The report examines the problem of authentication token abuse in Microsoft Entra ID, showing that even in the absence of obvious signs of compromise, an attacker can silently maintain access using stolen refresh tokens and obtain new access tokens. 🛑AI Guard: Securing LLMs, Agents, and AI Applications at Uber Scale. The authors present the AI Guard system — Uber's approach to securing AI applications, which includes input and output controls, security policies, monitoring, and protection mechanisms across the entire infrastructure. 🛑Inside the Hunt for China's Typhoons: Disrupt, Deter, and Defend. The report analyzes cyber operations by Chinese groups, emphasizing that their key strength lies not in sophisticated malware, but in stealth, patience, and the use of legitimate tools to maintain long-term presence within infrastructure.

💬 Discuss