C6 · AI Security Engineering (core)
AI Detection, Incident Response, and Evidence
Detect misuse, contain authority, preserve evidence, and reconstruct AI-system incidents.
Proposed
0 of 6 modules accessible. Joining Skool does not unlock planned modules.
At a glance
- Audience
- Security practitioners taking responsibility for AI systems.
- Prerequisites
- C1–C5 (or L1)
- Log analysis and incident-response fundamentals
- Planned duration
- 20 hours. Assumes the entry requirements are met. Planning estimate, not a measured completion time. Per module: 2 to 4 hours, plus practice.
- Assessment
- Incident replay, reconstruction and rubric.
- What is free
- The free foundation module listed on each course page, as it releases, plus community access.
Job tasks this course prepares you for
- Instrument AI services
- Detect misuse
- Contain agent authority
- Preserve evidence
- Reconstruct incidents
- Restore service
Learning outcomes
- Correlate a user request with agent identity, authorization decision, tool action, and resulting side effect.
- Detect all designated high-severity seeded incidents in the supplied replay, with false positives reported.
- Contain a compromised agent without destroying the specified evidence.
- Produce a timeline that separates observed facts, inference, and unresolved gaps.
Modules
00Evidence requirementsProposed
Topics
- identity, authorization decision, tool arguments and results, retrieval references, configuration versions.
01Privacy-conscious telemetryProposed
Topics
- redaction, access, retention, time synchronization, correlation, missing evidence.
02Detection engineeringProposed
Topics
- unusual authority, sensitive tool actions, repeated denials, egress, loops, spend anomalies.
03ContainmentProposed
Topics
- token revocation, tool disabling, workflow suspension, memory preservation, business coordination.
04InvestigationProposed
Topics
- timelines, provenance, entry point, actions, affected records, alternative explanations.
05RecoveryProposed
Topics
- restore approved state, validate controls, retain evidence, share lessons.
Labs and deliverables
- Investigate a compromised agent. Detection logic, evidence manifest, containment log, timeline, and incident report.
Labs run on SIFT, the forensics workstation Rob T. Lee created, on your own machine. Each module states what it needs before you start. Each module confirms its setup requirements and any cloud or API costs before it opens. About the SIFT lab environment.
Assessment
Incident replay, reconstruction and rubric. A named assessor grades against the published rubric (rubric to be published before release).
Pass criteria
- Evidence integrity verified
- Critical authority revoked
- Recovery test succeeds
- Detection scoring includes false positives
- 80% rubric
Critical failures
- Unauthorized cross-tenant access
- Out-of-scope activity
- Uncontrolled destructive action
- Fabricated evidence
- Exposed credentials
What this course does not teach
- Does not promise access to a model's hidden reasoning; traces are not treated as complete forensic truth.