Learning tracks by role
AI learning track: Security, governance and compliance
For security, risk and compliance teams reviewing AI data access, actions and accountability.
After this track: You can review an agent’s access and oversight, identify needed audit evidence and scope adversarial testing.
Follow these 9 steps in order, or return to the one you need for your current work. Reading times are estimates; listening times cover the full episode.
-
AI Security
Start with the AI security overview to map model and agent threats onto your existing security responsibilities.
-
Prompt Injection
Understand how untrusted content can become an instruction before assessing an agent’s access to documents and tools.
-
How do you give an AI agent an identity and permissions?
Bound what the agent can do even when instructions are manipulated before granting it production credentials.
-
Data Sovereignty
Identify where data may be processed and held before selecting the service that will receive it.
-
How do you decide what information an AI assistant may receive?
Translate that data boundary into a sharing policy employees can apply to real records.
-
How do you govern AI agents in an enterprise?
Assign owners and intervention rules once the system’s data and action boundaries are explicit.
-
Audit Trail
Specify the records needed to reconstruct an agent’s actions before relying on oversight that cannot be evidenced.
-
AI Red Teaming
Add adversarial testing to ordinary quality checks so the review also asks how the system can be made to fail.
-
Microsoft: planning red teaming for LLMs and their applications
Use a concrete planning guide to assign testers, choose application risks and record findings for follow-up.
Choose another role or continue through the five-level pathway.