CF05-016: AI Data Secure Disposal

AI Control Library / CF-05: Data Governance & Privacy

Designation: Core | Control types represented: Operational and Technical

Control requirement

Sanitize or otherwise securely dispose of AI data and storage media when information is no longer authorized for retention, using methods appropriate to media, service capabilities and sensitivity. Verify completion across the defined disposal boundary.

Why this matters

An application delete command may not sanitize retired media or remove exported copies. Cryptographic erasure can support sanitization only when key scope, encryption coverage and recoverable copies meet the method’s conditions.

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

Retired AI storage, artifacts containing protected data, temporary media, exports and supplier-held resources requiring secure disposal.

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 data owner authorizes disposal; storage and asset teams perform it with security and supplier coordination.

Implementation guidance

  1. Identify disposal targets, sensitivity, media type, relevant copies and any authorized preservation requirement.
  2. Select an approved sanitization method based on the medium and threat, checking vendor capabilities and validation evidence.
  3. Control custody through disposal; for cryptographic erasure assess keys, backups, plaintext remnants and shared-key consequences before execution.
  4. Verify and record the completed method and affected assets, addressing failed sanitization before reuse or release.

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
Trace a retired asset Approved disposal and custody evidence identify the relevant media or service scope.
Inspect a cryptographic-erasure claim Encryption coverage and all required key copies support the claimed result.
Test failure handling Unverified media is not released as successfully sanitized.

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

  • Disposal authorization and asset scope
  • Method and vendor validation
  • Custody and completion records
  • Verification and failure decisions

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 incomplete sanitization or uncertain custody to security and the asset owner. Quarantine the affected resource or use an approved alternative method before reuse or transfer.

Example in practice

Hypothetical example: A project retires a local model-evaluation disk. The team verifies sanitization of the disk and separately removes cloud exports instead of treating the local action as disposal of every copy.

Related controls

Framework mappings

  • CSA AICM (DSP-02)
  • NIST AI RMF (MANAGE 4.1)
  • EU AI Act (GDPR Art. 17)
  • 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

Scroll to Top