IriusRisk - Threat Modeling Agentic AI (IR-TM2AI)

Build fearless agentic AI. This course turns buzzwords into blueprints, clarifying what agents are, mapping their "brains", tamming their architecture, and shipping automation that’s powerful, auditable, and safe from day one using IriusRisk,

  • Course duration

    3 hours

  • Chapters

    4

  • Difficulty

    Advanced

  • CPE credits

    3

  • Cost

    Free of charge

  • Certification

    Included

  • Badge

    Included

    Learning Objective

    • Define agentic AI (vs. stateless LLMs) and why it changes security.
    • Model the architecture with layers & Trust Zones and draw a clear DFD.
    • Identify & map agent-specific threats to components and data flows.
    • Propose and prioritize controls for secure-by-design operation (least privilege, policy/safety gate, provenance, audit).
    • Reporting progress with Threat Modeling Agentic AI initiatives.

    Target Audience

    • Security/AppSec & AI Security: architects, red/blue teams securing agentic systems.
    • ML/AI & MLOps: engineers building/operating agents, RAG, tooling.
    • DevOps & Cloud: integrating agent actions into production workflows.
    • Software/Systems Architects: designing high-impact, automated services.
    • Product & Eng Managers; GRC/Compliance: owning risk, controls, and auditability.

    Course Master

    Agentic AI marks a new frontier, systems that act, decide, and shape outcomes. Threat modeling is our compass, ensuring this power is guided with intent, not left to drift into unintended consequences.
    George Makrodimitris
    Head of Software Adoption / Course Master