The Threat Modeling For Agentic Ai Masterclass

Master How To Threat Model Agentic AI systems Using The MAESTRO Framework and OWASP Best Practises

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

Requirements

Basic knowledge of AI / GenAI

Good knowledge of Cybersecurity

Desire to Learn

Description

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 AI Threats 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 LearnThe core principles and architecture of Agentic AI and multi-agent systemsHow to apply the MAESTRO framework for layered threat modelingThe OWASP Agentic AI (T1–T15) threat taxonomy and how it applies in practiceAttack techniques against Agentic AI, including tool misuse, goal manipulation, memory poisoning, and human-in-the-loop exploitationHow to conduct risk assessments and threat model these systemsStrategies for building mitigation plans that cover both AI-specific and cross-layer security risksCourse OutlineIntroduction to Agentic AI SystemsWhat are Agentic AI and multi-agent systems?How do they differ from traditional and generative AI?Why security in Agentic AI is non-negotiableThreats in Agentic AI SystemsOverview of the Agentic AI risk landscapeThe ASI Threat Model (T1–T15) explainedThe MAESTRO FrameworkFoundation Model risksData Operations threats (RAG poisoning, communication attacks)Agent Framework misuse (tool abuse, intent breaking)Deployment, Observability, Ecosystem securityApplying Agentic AI Threat Modeling in PracticeStep-by-step threat modeling with MAESTROThree Case Studies with eternal resources (Notion Templates)Mitigation and GovernanceDesigning secure architectures for Agentic AILayer-specific and cross-layer mitigationsWho Should Take This CourseThis course is ideal for individuals seeking to secure the next generation of AI systems, including:AI engineers and architectsCybersecurity professionalsData scientists and AI governance specialistsIT managers and risk professionalsBusiness leaders evaluating Agentic AI adoptionPre-requisitesA basic understanding of AI and cybersecurity is recommendedInstructorTaimur 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.

Overview

Section 1: Introduction

Lecture 1 Introduction

Section 2: Agentic AI Foundations

Lecture 2 What is Agentic AI

Lecture 3 Agentic AI Architecture

Lecture 4 Demo - Agentic AI

Lecture 5 The Model Context Protocol

Lecture 6 Demo - MCP

Section 3: Agentic AI Threats and Risks

Lecture 7 Understanding Agentic AI Threats

Lecture 8 T1 - Memory Poisoning

Lecture 9 T2 - Tool Misuse

Lecture 10 T2 - Tool Misuse ( Assessing MCP Servers )

Lecture 11 T3 - Privilege Compromise

Lecture 12 T4 - Resource Overload

Lecture 13 T5 - Cascading Hallucinations

Lecture 14 T6 - Intent Breaking

Lecture 15 T7 - Misalignment

Lecture 16 T8 - Repudiation

Lecture 17 T9 - Identity Spoofing

Lecture 18 T10 - Overwhelming Human in the Loop

Lecture 19 T11 - Remote Code Execution

Lecture 20 T12 - Agent Communication

Lecture 21 T13 - Rogue Agent

Lecture 22 T14 - Human Attacks on Multi-Agentic Systems

Lecture 23 T15 - Human Manipulation

Section 4: Threat Modeling Agentic AI

Lecture 24 Threat Modeling Agentic AI Part 1

Lecture 25 Threat Modeling Agentic AI Part 2

Lecture 26 Case Study 1 - Part 1

Lecture 27 Case Study 1 - Part 2

Lecture 28 Case Study 1 - Part 3

Lecture 29 Case Study 1 - Notion Template

Lecture 30 Case Study 2 - Part 1

Lecture 31 Case Study 2 - Part 2

Lecture 32 Case Study 2 - Notion Template

Lecture 33 Case Study 3

Lecture 34 Case Study 3 - Notion Template

Section 5: Designing Secure Agentic Systems

Lecture 35 Secure Agentic AI Design Patterns - Part 1

Lecture 36 Secure Agentic AI Design Patterns - Part 2

Section 6: Conclusion

Lecture 37 The Way Forward

Cybersecurity Professionals,AI Security Professionals,Agentic AI Professionals,GenAI Professionals,CISOs, CTOs, CROs,Threat Modeling Experts,IT Risk Management Professionals,DevSecOps Engineers,Anyone wanting to learn this topic.