What you'll learn:
- Understand Agentic AI Architecture and Components
- Analyze and mitigate threats unique to Agentic AI based on OWASP
- Apply structured threat modeling techniques such as MAESTRO
- Design secure agentic systems
Agentic AI represents the next evolution of artificial intelligence—systems that can autonomously plan, make decisions, and execute actions with minimal human input. These multi-agent ecosystems are transforming industries, but they also introduce new security risks that extend far beyond traditional cybersecurity concerns.
The "Threat Modeling Agentic AI Systems Masterclass" is a practical, hands-on course designed to teach you how to identify, analyze, and mitigate threats in autonomous AI systems using structured frameworks like MAESTRO and the OWASP Agentic AIThreats and Mitigations Guide.
This course goes beyond theory—by walking through case studies, real-world scenarios, and layered defenses, you will learn how to systematically map threats to risks (T1–T15), evaluate their likelihood and impact, and design effective mitigations.
What You Will Learn
The core principles and architecture of Agentic AI and multi-agent systems
How to apply the MAESTRO framework for layered threat modeling
The OWASP Agentic AI (T1–T15) threat taxonomy and how it applies in practice
Attack techniques against Agentic AI, including tool misuse, goal manipulation, memory poisoning, and human-in-the-loop exploitation
How to conduct risk assessments and threat model these systems
Strategies for building mitigation plans that cover both AI-specific and cross-layer security risks
Course Outline
Introduction to Agentic AI Systems
What are Agentic AI and multi-agent systems?
How do they differ from traditional and generative AI?
Why security in Agentic AI is non-negotiable
Threats in Agentic AI Systems
Overview of the Agentic AI risk landscape
The ASI Threat Model (T1–T15) explained
The MAESTRO Framework
Foundation Model risks
Data Operations threats (RAG poisoning, communication attacks)
Agent Framework misuse (tool abuse, intent breaking)
Deployment, Observability, Ecosystem security
Applying Agentic AIThreat Modeling in Practice
Step-by-step threat modeling with MAESTRO
Three Case Studies with eternal resources (Notion Templates)
Mitigation and Governance
Designing secure architectures for Agentic AI
Layer-specific and cross-layer mitigations
Who Should Take This Course
This course is ideal for individuals seeking to secure the next generation of AI systems, including:
AI engineers and architects
Cybersecurity professionals
Data scientists and AI governance specialists
IT managers and risk professionals
Business leaders evaluating Agentic AI adoption
Pre-requisites
A basic understanding of AI and cybersecurity is recommended
Instructor
Taimur Ijlal is a multi-award-winning cybersecurity leader with over 20 years of global experience in cyber risk management, AI security, and IT governance. Recognized as CISO of the Year and one of the Top 30 CISOs worldwide, Taimur’s work has been featured in ISACA Journal, CIO Magazine Middle East, and multiple AI security publications.
He has trained thousands of students worldwide through his Udemy courses, and his books on AI Security and Cloud Computing have ranked as #1 New Releases on Amazon.
