Post Content
Welcome back to AI Red Teaming 101!
In this episode, Gary Lopez, Principal Offensive AI Scientist at Microsoft, walks us through PyRIT, the open-source tool he created to help automate and scale adversarial testing of generative AI systems, now supported by the Microsoft AI Red Team.
Gary walks through PyRIT’s architecture, core components (like datasets, orchestrators, converters, and scoring engines), and demonstrates how it can be used to automate attacks on models like Gandalf. Whether you’re a security researcher or ML engineer, PyRIT helps you move from manual testing to repeatable, scalable red teaming workflows.
What You’ll Learn:
How PyRIT automates red teaming tasks for generative AI
Key components of PyRIT’s architecture and how they work together
How to use PyRIT to simulate real-world attacks and evaluate model robustness
✅ Chapters:
00:00 – Welcome & episode overview
00:30 – Why we built PyRIT
01:00 – What PyRIT automates
01:40 – Four reasons to use PyRIT
03:00 – PyRIT’s architecture overview
04:00 – Targets, scoring engines, and memory
06:30 – Demo: PyRIT vs. Gandalf
07:30 – Final thoughts & where to learn more
✅ Links & Resources:
AI Red Teaming 101 Episodes: aka.ms/airt101
AI Red Teaming 101 Labs & Tools: aka.ms/airtlabs
Microsoft AI Red Team Overview: aka.ms/airedteam
PyRIT GitHub Repository: https://github.com/Azure/PyRIT
✅ Speakers:
Amanda Minnich – Principal Research Manager, Microsoft AI Red Team
LinkedIn: https://www.linkedin.com/in/amandajeanminnich/
Webpage: https://www.amandaminnich.info/
Gary Lopez – Principal Offensive AI Scientist, ADAPT
LinkedIn: https://www.linkedin.com/in/gary-lopez/
#AIRedTeam #AIRT #Microsoft #AI #AISecurity #AIRedTeaming #GenerativeAI #Cybersecurity #InfoSec #cybersecurityawareness #PyRIT Read More Microsoft Developer
