AI Control Library / CF-12: Security & Abuse Resistance
Designation: Core | Control types represented: Operational and Technical
Control requirement
Grant each AI-related identity only the permissions necessary for its approved duties. Enforce access at the underlying resource and review cumulative privileges across connected systems.
Why this matters
A narrow user interface can conceal broad service-account rights. If the backend can read every tenant or change production settings, model instructions alone cannot contain misuse.
How the control types work together
Operational controls define scope, authority and the actions required when conditions are not met. Technical controls produce evidence or enforce the relevant boundaries. Both must apply to the same deployment and operating conditions.
When it applies
Apply to human users, agents, workloads, integrations and administrators across the full AI service and data lifecycle.
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
Resource owners approve access; identity teams enforce roles; service owners justify machine permissions.
Implementation guidance
- Map roles and tasks to required resources and actions.
- Remove broad defaults and separate read, write and administrative permissions.
- Constrain service credentials by resource, tenant and environment.
- Review effective privileges and test denied actions.
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 |
|---|---|
| Out-of-role action | Underlying service denies access. |
| Unused permission review | Unnecessary grants are removed. |
| Cross-tenant request | Tenant boundary remains enforced. |
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
- Role matrix
- Approval records
- Effective permission reviews
- Denial tests
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
Revoke excess privileges, investigate relevant use and reauthorize only the minimum access supported by current duties.
Example in practice
Hypothetical example: A summarization service receives read access to one approved document collection. Its credential cannot delete documents or retrieve another department’s files.
Related controls
- CF12-004: AI Knowledge Base & RAG Access Control
- CF12-006: AI Agent Access Restriction
- CF12-S12: AI System Authorization Mechanisms
Framework mappings
- CSA AICM (IAM-05, IAM-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