CWTA-14
TA-14 AI GOVERNANCE LIBRARY
AI Governance Crosswalk Engine
Compare governance concepts across major authorities, identify meaningful overlap and divergence, preserve interpretation boundaries, and translate source relationships into TA-14 admissible execution routes.
STEP 01 · SELECT THE COMPARISON
Choose the authorities and concepts to examine.
Crosswalks identify conceptual relationships. They do not establish legal equivalence, certification, conformity, or interchangeable compliance.
CROSSWALK SUMMARY
EU AI Act compared with ISO/IEC 42001.
STEP 02 · INSPECT THE RELATIONSHIPS
Concept-level governance mapping.
Each concept includes source support status, TA-14 interpretation, recommended evidence, and the execution route needed to move from guidance to controlled action.
| Governance Concept | EU AI Act | ISO/IEC 42001 | TA-14 Interpretation | Execution Route |
|---|---|---|---|---|
01 RiskRisk Management The structured identification, analysis, evaluation, treatment, monitoring, and review of AI-related risk. | Supported | Supported | Risk claims are converted into evidence-bound admissibility gates, review thresholds, and controlled execution conditions. Risk RegisterRisk AssessmentRisk Treatment DecisionResidual Risk RecordMonitoring Criteria | TA-14 Risk and Evidence RouteBuild route → |
02 AuthorityHuman Oversight The allocation of meaningful human authority, intervention capability, escalation responsibility, and review control. | Supported | Supported | Human oversight is mapped to authority validation, escalation rights, intervention evidence, and preserved decision accountability. Authority RecordOversight PlanEscalation ProcedureIntervention LogReview Decision | TA-14 Authority and Oversight RouteBuild route → |
03 DocumentationTechnical Documentation The structured documentation necessary to understand a system, its intended purpose, design, limitations, controls, and performance. | Supported | Supported | Documentation becomes governed evidence only when source identity, version, continuity, authority, and execution relevance are preserved. System DescriptionModel DocumentationData DocumentationVersion RecordChange History | TA-14 Governed Documentation RouteBuild route → |
04 DataData Governance The governance of data quality, provenance, relevance, representativeness, access, processing, retention, and use. | Supported | Supported | Data cannot support execution unless its identity, provenance, integrity, authorized use, and relationship to the decision are established. Data InventoryData Provenance RecordQuality AssessmentAccess RecordRetention Policy | TA-14 Data Admissibility RouteBuild route → |
05 TrustworthinessTransparency The disclosure and communication necessary for affected parties, operators, reviewers, and authorities to understand AI use and limitations. | Supported | Supported | Transparency is preserved as attributable disclosure evidence tied to system identity, operating context, decision boundaries, and affected parties. Transparency NoticeUser DisclosureSystem Limitation RecordDecision ExplanationAffected-Party Communication | TA-14 Transparency Evidence RouteBuild route → |
06 AuthorityAccountability The assignment and preservation of responsibility for governance decisions, operational actions, review, and outcomes. | Supported | Supported | Accountability requires attributable actors, validated authority, bound decisions, execution receipts, and preserved outcome evidence. Responsibility MatrixAuthority AssignmentApproval RecordExecution ReceiptOutcome Review | TA-14 Accountability RouteBuild route → |
07 AssuranceTesting and Validation The structured examination of system performance, reliability, safety, limitations, and control effectiveness. | Supported | Supported | Testing evidence is evaluated for method, scope, authority, continuity, limitations, and relevance before it can support execution. Test PlanValidation ResultsPerformance ThresholdsFailure RecordIndependent Review | TA-14 Testing and Assurance RouteBuild route → |
08 EvidenceLogging and Traceability The preservation of records capable of reconstructing system activity, decisions, changes, interventions, and outcomes. | Supported | Supported | Logs become admissible records only when identity, time, sequence, integrity, custody, continuity, and decision relevance are preserved. System LogDecision RecordChange LogChain-of-Custody RecordReplay Package | TA-14 Traceability and Replay RouteBuild route → |
09 OperationsIncident Management The detection, reporting, containment, investigation, correction, and review of AI-related incidents and failures. | Supported | Supported | Incident evidence can trigger HOLD, DENY, ESCALATE, corrective action, revalidation, and preservation of the post-incident outcome. Incident ReportContainment RecordRoot-Cause AnalysisCorrective ActionRevalidation Record | TA-14 Incident Revalidation RouteBuild route → |
10 LifecyclePost-Market Monitoring The continuous collection and evaluation of operational evidence after deployment. | Supported | Supported | New operational evidence can invalidate prior assumptions and trigger runtime revalidation before further execution is permitted. Monitoring PlanOperational MetricsDrift RecordComplaint RecordRevalidation Decision | TA-14 Continuous Admissibility RouteBuild route → |
11 RiskImpact Assessment The evaluation of foreseeable effects on people, rights, safety, organizations, infrastructure, and society. | Supported | Supported | Impact claims are separated from proof and bound to declared scope, affected parties, evidence quality, uncertainty, and authorized review. Impact AssessmentAffected-Party AnalysisRights ReviewSafety AssessmentMitigation Record | TA-14 Impact Admissibility RouteBuild route → |
12 LifecycleChange Management The governance of system, model, data, configuration, purpose, ownership, and operating-environment changes. | Partial | Supported | A material change breaks prior continuity until the new state, authority, evidence, controls, and execution conditions are revalidated. Change RequestImpact ReviewVersion RecordApproval RecordPost-Change Validation | TA-14 Change Revalidation RouteBuild route → |
STATUS INTERPRETATION
Read the crosswalk without overstating it.
The source expressly addresses the concept at a meaningful level.
The source addresses part of the concept, uses a different scope, or requires interpretation.
The concept is not clearly stated in the selected source and should not be inferred as equivalent.
The relationship cannot be responsibly stated without source-level and contextual review.
XBCrosswalk boundary
SOURCE RELATIONSHIP BOUNDARY
Similar governance concepts do not create equivalent authority.
A crosswalk can identify relationships, shared themes, missing concepts, and implementation opportunities. It cannot convert a voluntary framework into law, replace a standard with a principle, establish conformity, or prove that satisfying one source satisfies another.