Definition: Standards mapping documents how Attested AI artifacts provide evidence for governance frameworks. Mechanism: Each artifact field maps to specific control requirements. Value: Auditors and governance teams can trace evidence to standards. Output: Crosswalk tables linking artifacts to controls.

Audit Documentation Reference

Standards Mapping

How governance evidence artifacts may support audit workflows for industry standards.

Important: Attested AI artifacts may provide documentation for audit workflows related to the standards below. This documentation describes how artifacts can provide evidence for specific controls. Actual compliance depends on your implementation, operational context, and auditor assessment. This is not a certification or compliance guarantee.

NIST AI RMF 1.0

AI Risk Management Framework (NIST AI 100-1)

FunctionCategorySupporting Evidence
GOVERNGV-1: Governance policiesPolicy Artifact (policy_version, enforcement_mapping)
GOVERNGV-3: Accountability structuresEnforcement Receipt (signer.key_id, decision.action)
MAPMP-2: AI system categorizationPolicy Artifact (subject.subject_type, measurement_set)
MEASUREMS-1: Metrics and methodsEnforcement Receipt (measurement.composite_hash, mismatched_paths)
MEASUREMS-2: Continuous monitoringContinuity Chain (hash-linked receipts, chain_head)
MANAGEMG-2: Risk responseEnforcement Receipt (decision.action: KILL/QUARANTINE)
MANAGEMG-4: Documentation and reportingEvidence Bundle (bundle_manifest, offline verification)

NIST SP 800-53 Rev. 5

Security and Privacy Controls for Information Systems

FamilyControlSupporting Evidence
AUAU-2: Event loggingEnforcement Receipt (event_type, timestamp, decision)
AUAU-9: Protection of audit infoReceipt signatures (Ed25519), hash chain integrity
AUAU-10: Non-repudiationCryptographic signatures on all artifacts (issuer.signature, signer.signature)
CMCM-2: Baseline configurationPolicy Artifact (integrity_policy, measurement_set baselines)
CMCM-3: Configuration change controlDrift Detection (DRIFT_DETECTED receipts, policy versioning)
SISI-4: System monitoringContinuous measurement (MEASUREMENT_OK receipts, schedule.interval_ms)
SISI-7: Software integritySubject manifest (SHA-256 digests), composite_subject_hash
SASA-10: Developer config managementPolicy Artifact (config_digest, sbom_digest references)

ISO/IEC 42001

Artificial Intelligence Management System (AIMS)

ClauseRequirementSupporting Evidence
5.2AI PolicyPolicy Artifact (versioned, signed governance document)
6.1Risk assessmentPolicy Artifact (drift_rules, enforcement_mapping define risk responses)
7.5Documented informationEvidence Bundle (deterministic, versioned, offline-verifiable)
8.2AI system lifecycleContinuity Chain (complete lifecycle receipts from POLICY_LOADED to BUNDLE_EXPORTED)
9.1Monitoring, measurement, analysisEnforcement Receipts (measurement.composite_hash, continuous integrity checks)
9.2Internal auditOffline Verifier (deterministic audit without network dependency)
10.1Nonconformity and corrective actionDRIFT_DETECTED + ENFORCED receipts (documented nonconformity and response)

Artifact-to-Control Summary

Policy Artifact provides evidence for:

  • Governance policies (GV-1, 5.2)
  • Baseline configurations (CM-2)
  • Risk assessment responses (6.1)
  • System categorization (MP-2)

Enforcement Receipt provides evidence for:

  • Event logging (AU-2)
  • Non-repudiation (AU-10)
  • Configuration change control (CM-3)
  • System monitoring (SI-4)

Continuity Chain provides evidence for:

  • Audit info protection (AU-9)
  • Continuous monitoring (MS-2)
  • AI system lifecycle (8.2)

Evidence Bundle provides evidence for:

  • Documentation and reporting (MG-4)
  • Documented information (7.5)
  • Internal audit (9.2)