ARCS/Crosswalks/ARCS / Singapore AI Verify Crosswalk
ARCS / Singapore AI Verify Crosswalk
Source-element links point to official public source materials where stable public URLs are available. For standards or criteria that do not expose freely accessible clause-level text, links point to the official product, browsing, or reference page. The mapping remains informative.
Overview
AI Verify is an AI governance testing framework that helps organizations assess responsible AI implementation against internationally recognized AI governance principles, including process and evidence-based assessment.
ARCS is a separate lifecycle-governance standard for records created during AI system use. This crosswalk identifies where ARCS supports governance evidence, test outputs, assessment reports, and custody surfaces created by AI assurance activity.
Interpretive status
This instrument is an informative crosswalk. It does not replace AI Verify, does not assess whether an AI system satisfies AI Verify, and does not convert ARCS conformance into responsible AI assurance.
Framework scope
AI Verify addresses responsible AI implementation through governance principles, processes, evidence, technical tests, and testing reports. The framework has been updated to support both traditional AI and generative AI use cases.
ARCS relevance
AI Verify can create or depend on assessment records, process evidence, test outputs, report artifacts, and stakeholder-facing assurance materials. ARCS governs the lifecycle and custody posture of those records after they are created.
Selected mappings
Table A maps selected AI Verify themes to ARCS control families. The mapping is directional and explanatory, not a principle-by-principle equivalence claim.
| AI Verify theme | Reference | ARCS families | Crosswalk note |
|---|---|---|---|
| Governance principles and desired outcomes | AI Verify Testing Framework | ARCS-OPB, ARCS-TAX, ARCS-VER | AI Verify organizes responsible AI assessment around governance principles and desired outcomes. ARCS supplies a record-governance layer for the records that evidence those outcomes, including operator responsibilities, record classes, and verification posture. |
| Process implementation and documentary evidence | AI Verify process checks | ARCS-LIF, ARCS-PV, ARCS-VER | AI Verify validates processes through documentary evidence. ARCS governs the lifecycle, preservation state, and verification posture of the documentary evidence created by AI governance processes. |
| Technical testing and generated reports | AI Verify technical testing tools and reports | ARCS-LIF, ARCS-TAX, ARCS-PUB, ARCS-VER | Technical testing can generate reports, outputs, benchmark results, and supporting artifacts. ARCS classifies those records, defines lifecycle treatment, governs publication or sharing boundaries, and supports later verification of report provenance. |
| Transparency and explainability | AI Verify transparency and explainability principles | ARCS-TAX, ARCS-PUB, ARCS-VER | Transparency and explainability efforts often create explanatory records, model cards, user notices, assessment outputs, and disclosure materials. ARCS clarifies the lifecycle and publication status of those records without treating disclosure alone as lifecycle governance. |
| Fairness, robustness, safety, and security testing | AI Verify technical and governance principles | ARCS-VER, ARCS-PV, ARCS-LIF | Testing for fairness, robustness, safety, and security creates evidence that may be reused in assurance, procurement, audit, or incident review. ARCS governs preservation, lifecycle state, and verification of those evidence records. |
| Human agency and accountability | AI Verify governance principles | ARCS-OPB, ARCS-DEL, ARCS-AGT | AI Verify addresses human agency and accountability at the governance layer. ARCS maps the record consequences of delegated decisions, agent tool use, approval chains, human intervention, and operator boundary allocation. |
| Data governance and record provenance | AI Verify governance principles | ARCS-CUS, ARCS-TAX, ARCS-VER | AI Verify addresses data governance as part of responsible AI implementation. ARCS focuses on provenance and custody for interaction records, including where records are generated, retained, transformed, exported, or verified. |
| Traditional AI and generative AI coverage | AI Verify traditional and generative AI framework coverage | ARCS-LIF, ARCS-AGT, ARCS-DEL, ARCS-NCR | AI Verify has been updated to address both traditional and generative AI use cases. ARCS contributes generative-AI-specific lifecycle treatment for prompts, outputs, memory, tool use, delegation, and non-creation claims. |
| Assurance communication and report sharing | AI Verify reports and stakeholder transparency | ARCS-PUB, ARCS-PV, ARCS-CUS | AI Verify testing reports may be shared with stakeholders. ARCS governs the publish boundary, preservation consequences, and custody posture of reports and supporting records that leave the original testing environment. |
Outside scope
ARCS governs several record-lifecycle domains that remain distinct from AI Verify testing and assurance activity:
Model-performance testing
ARCS does not perform or replace technical testing of fairness, explainability, robustness, safety, or model quality. It governs the records produced by those tests.
Responsible AI certification or endorsement
ARCS Section 16; ARCS-VER
ARCS conformance does not establish AI Verify assessment success, responsible AI certification, legal compliance, or endorsement by Singapore authorities or the AI Verify Foundation.
General AI governance program design
ARCS does not replace organizational AI governance frameworks. It supplies lifecycle, custody, classification, preservation, and verification treatment for records created by those frameworks.