Security Architects Track¶
By the end of this track you will be able to¶
- Map multi-agent failure modes to the AIRS three-layer defence architecture
- Identify where epistemic integrity controls must sit in a pipeline
- Select MASO controls that address reasoning-basis failures, not just output failures
- Design verification evidence that proves controls are working
- Make informed intervene/allow decisions on ambiguous agent chain signals
Your thread¶
As a security architect, your job is to design systems that catch failures before they cause damage, and to do that, you need to understand the threat model deeply enough to know where to place controls.
This track takes you through:
| Module | Focus |
|---|---|
| 1. What Goes Wrong | Threat modelling multi-agent systems: where the attack surface actually is |
| 2. Why Controls Miss It | Architectural gaps in standard AI security patterns |
| 3. Epistemic Integrity | The core concept and how to design for it |
| 4. MASO Controls | Control domains that address the gap, and how they integrate |
| 5. Verification & Evidence | How to know your controls are actually working |
| Decision Exercise | Ambiguous signals: intervene or allow? |