AI Control Library / CF-05: Data Governance & Privacy
Designation: Supplemental | Control types represented: Operational
Control requirement
Establish an encryption and key management policy for AI that defines approved protection, ownership, access, lifecycle, recovery and compromise handling. Align cryptographic requirements with data sensitivity and service dependencies.
Why this matters
An encryption requirement is incomplete without a plan for the keys. Unclear lifecycle and recovery rules can expose information or make retained data permanently unavailable.
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
Cryptographic services protecting AI information and supporting infrastructure, including provider-managed and customer-managed arrangements.
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 security cryptography owner maintains the policy; data and service owners define use and continuity requirements.
Implementation guidance
- Define covered data and protection states, approved cryptographic services and responsibility for provider and customer components.
- Set key purpose, access, lifecycle, cryptoperiod and assurance requirements proportionate to the threat and information lifetime.
- Define generation, rotation, revocation, destruction, archival, recovery and compromise processes with appropriate authority.
- Integrate requirements into acquisition and change review, testing whether services can meet them without undocumented exceptions.
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 |
|---|---|
| Inspect an AI encryption deployment | Its protection and key lifecycle follow an approved standard. |
| Trace a lifecycle exception | Owner, justification and limits are documented. |
| Review recovery and compromise procedures | Both preserve the required security and availability conditions. |
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
- Encryption and key policy
- Approved service and lifecycle standards
- Responsibility and exception records
- Policy implementation reviews
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
Escalate unsupported cryptographic requirements or unmanaged key services to security and business owners. Restrict sensitive processing or establish an approved alternative.
Example in practice
Hypothetical example: A service encrypts backups but has no identified key recovery owner. Policy review closes the gap before those backups are accepted as a reliable recovery capability.
Related controls
- CF05-017: AI Data Encryption at Rest & in Transit
- CF05-S02: AI Encryption Roles & Responsibilities
- CF05-S05: AI Encryption Risk Management
Framework mappings
- CSA AICM (CEK-01)
- 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 Data Governance & Privacy
Version 1.0