AI Control Library / CF-12: Security & Abuse Resistance
Designation: Supplemental | Control types represented: Operational
Control requirement
Maintain accurate network and connectivity documentation for AI infrastructure, including trust boundaries, ingress, egress, administrative paths and supplier connections. Reconcile documentation with deployed configuration.
Why this matters
Incident responders cannot reliably contain a service when its dependencies and routes are unknown. Outdated diagrams also obscure unauthorized exposure and shared responsibilities.
How the control types work together
Operational controls establish the assigned responsibilities, review processes and response decisions described here. Their effectiveness depends on performing the work and using evidence to guide action, rather than approving a document alone.
When it applies
Apply to physical, virtual and service-level connections across AI applications, data systems, providers and management infrastructure.
Define the assessed boundary, dependencies and evidence limitations. Tailor depth and review frequency to potential impact, and document why a particular safeguard or exclusion is appropriate to the actual deployment.
Ownership
The network architecture owner maintains documentation; platform teams report changes; service owners verify dependencies.
Implementation guidance
- Record components, endpoints, protocols, identities and data flows.
- Mark trust boundaries and management or external connections.
- Update documentation as part of controlled infrastructure changes.
- Reconcile diagrams and inventories with observed configuration and traffic.
How to verify it
Use representative records and authorized controlled tests. Identify the relevant version, operating conditions and expected result before testing, and retain enough evidence to support the conclusion.
| Test | Expected result |
|---|---|
| Deployed endpoint sample | Documented owner and purpose match reality. |
| New connection introduced | Documentation updates with the approved change. |
| Incident exercise | Responders identify isolation points and dependencies. |
Record a finding when required results cannot be demonstrated or practice differs materially from the approved conditions. Missing evidence limits the conclusion; it should not be treated as a successful test.
Evidence to retain
- Network diagrams
- Connection inventory
- Reconciliation results
- Change records
Protect evidence according to its sensitivity, access needs and retention rules. Avoid unnecessary copies of personal information, secrets or confidential material, while preserving the records needed for authorized review.
If the control fails
Investigate undocumented connections, update affected response plans and restrict unexplained paths according to exposure.
Example in practice
Hypothetical example: An incident exercise reveals that a retrieval service reaches an unrecorded supplier endpoint. The team identifies its owner and updates both connectivity rules and the response map.
Related controls
- CF12-S15: AI Infrastructure Network Security
- CF12-S18: AI Infrastructure Segmentation
- CF12-S21: AI Infrastructure Network Defense
Framework mappings
- CSA AICM (I&S-08)
- NIST AI RMF (MAP 1.1)
- ISO 42001 (Cl. 8.2)
Review and maintenance
Review at the interval approved by the control owner and after relevant changes, incidents or failed assessments. Update affected procedures and configuration, address evidence gaps and verify corrective action before closing findings.
Back to Security & Abuse Resistance
Version 1.0