EN 18286:2025 — Quality management system for high-risk AI systems
EN 18286:2025 — Quality management system for high-risk AI systems
EN 18286:2025 — Quality management system for high-risk AI systems — 83% of in-scope requirements covered.
6 requirements · 6 in scope (4 enforced · 2 designed). The 83% is weighted over the in-scope base.
Source: EN 18286:2025 (E) — Clause 5.1 (General management responsibility for high-risk AI QMS) · License: CEN/CENELEC copyright — citation only
By category
| Category | Reqs | Enforced | Designed | Advisory | Deferred | Coverage |
|---|---|---|---|---|---|---|
| Clause 5 — Management responsibility (5.1 General) | 6 | 4 | 2 | 0 | 0 | 83% |
Every requirement → the KYE™ artefact that enforces it
| ID | Title | Status | KYE™ enforcement |
|---|---|---|---|
en-18286.5.1.a |
Top management shall establish the quality policy and measurable quality objectives, derived from the regulatory purpose (preventing harm to health, safety and fundamental rights), not internal efficiency | enforced | audit_events: kye.purpose.permission.v1, kye.risk.authority_register.v1, kye.compliance.attestation.v1engines: internal, internal, internalconstitution_refs: constitution/12-PURPOSE-PERMISSION.md |
en-18286.5.1.b |
Top management shall ensure the necessary resources for the QMS (Clause 7): infrastructure/compute, human capital with verified competence, sustainability, and storage for data lineage/traceability across the full retention period | designed | audit_events: kye.audit_retention_policy.v1, kye.compliance.attestation.v1constitution_refs: constitution/30-AUDIT-WORM-RETENTION.md |
en-18286.5.1.c |
Top management shall ensure other roles carry out their responsibilities effectively: clear accountability, human-oversight procedures, intervention thresholds, intervenability (a human reviewer with the authority and technical interface to OVERRIDE AI output), continuous competence monitoring, and mitigation of automation bias | enforced | audit_events: kye.purpose.admissibility.v1, kye.compliance.attestation.v1engines: internal, internalgovernedui_modules: kye.governedui.module.approval_queue.v1, kye.governedui.module.oversight_envelope.v1, kye.governedui.module.authority_drift.v1constitution_refs: constitution/36-GOVERNEDUI.md |
en-18286.5.1.d |
Top management shall integrate the QMS requirements into the provider's processes across the whole AI lifecycle (data management, pre-deployment modifications, change management), embedded into the existing sectoral QMS — not a separate binder | enforced | audit_events: kye.evidence.decision_map.v1, kye.compliance.attestation.v1engines: internalconstitution_refs: constitution/53-COHESION-CASCADE.md, constitution/34-RECONCILIATION-ENGINE.md |
en-18286.5.1.e |
Top management shall ensure the QMS achieves its intended results (management review; nonconformities trigger corrective action) | enforced | audit_events: kye.compliance.attestation.v1constitution_refs: constitution/34-RECONCILIATION-ENGINE.mdreconcilers: agent-manifest-alive, engine-manifest-alive |
en-18286.5.1.f |
Top management shall communicate the importance of the QMS and promote a responsible-AI culture, sustaining competence and culture amid staff turnover and drift | designed | audit_events: kye.compliance.attestation.v1constitution_refs: constitution/38-COMMS-RAIL.md, constitution/39-LEARN-RAIL.mdrails: 38 — Comms Rail · KYE Comms Engine™, 39 — Learn Rail · KYE Learn™ |