# WIA-ai-survival-2026 PHASE 4 — INTEGRATION Specification **Standard:** WIA-ai-survival-2026 **Phase:** 4 — INTEGRATION **Version:** 1.0 **Status:** Stable This document defines how an AI-survival operator integrates with the systems that surround the AI- governance lifecycle: the AI-system providers and deployers in the operator's value chain (cross- reference WIA-generative-ai + WIA-content-ai); the AI Safety Institute network (US AISI + UK AISI + EU AI Office + KAISI + Japan / Singapore / Canada); the international coordination bodies (UN AI Advisory Body, GPAI, OECD AI Network, UNESCO); the EU AI Office and the Member-State NCAs; the US sector regulators (FTC + EEOC + DOT NHTSA + FDA + CFPB); the KR AI 기본법 supervisory regime; the workforce- transition partners (educational institutions, public employment agencies, sectoral training funds, trade unions); the responsible-scaling-policy commitment network (Seoul AI Summit / AI Action Summit); the external auditor and ISO/IEC 27001 + ISO/IEC 42001 certification body; the dual-use export- control authority; and the long-term archive. References (CITATION-POLICY ALLOW only): - ISO/IEC 42001:2023 + 27001:2022 + 22989 + 23053 + 23894 + 24029 + TR 24027 + TR 24028 + 38507 - IETF RFC 8259 / 9457 / 8615 / 8288 / 9421 - ISO/IEC 17021-1:2015, ISO/IEC 17065:2012 - ISO 8601, ISO 17442 LEI - W3C Verifiable Credentials Data Model 2.0 - NIST AI RMF + NIST AI 600-1 + NIST IR 8332 - US AISI Consortium voluntary commitments + US EO 14110 + OMB M-24-10 - EU AI Act Articles 49 (registration), 51-55, 70 (governance), 71 (database), 72 (post-market), 73 (serious-incident), 74 (market surveillance), 86 (right to lodge complaint) - EU Code of Practice for GPAI Models - UN AI Advisory Body + UN GA Resolution 2024 + UN GDC AI provisions - International AI Safety Report 2025 - OECD AI Principles + UNESCO Ethics Recommendation 2021 - ETSI ISG SAI - IEC 62443 - KR AI 기본법 + KR AI 안전연구소 (KAISI) + KR PIPC + KR 대외무역법 --- ## §1 AI-System Provider and Deployer Integration The operator's integration with the upstream and downstream AI-system actors: - Provider integration — when the operator deploys a third-party AI system the upstream provider's WIA-generative-ai (or equivalent) record set feeds the operator's AI-system inventory. - Deployer integration — for high-risk AI systems the operator's deployer obligations under EU AI Act Article 26 are coordinated with the system's provider through the published instruction-for- use. - Distributor integration — distributor obligations under Article 24. - Importer integration — importer obligations under Article 23. ## §2 AI Safety Institute Network Integration The operator's AISI integration: - US AISI (NIST) — the US AI Safety Institute Consortium voluntary participation; pre-deployment testing memorandum-of-understanding for frontier models. - UK AISI — the UK AI Safety Institute pre- deployment evaluations. - EU AI Office — the European Commission's AI Office for GPAI oversight under AI Act Articles 64-70 + Code of Practice signature. - KR AI 안전연구소 (KAISI) — the Korean AI Safety Institute (under MSIT) for KR-jurisdiction pre- deployment evaluation. - Japan AISI / Singapore AI Verify Foundation / Canada AISI — bilateral and multilateral cooperation. - AISI international network — bilateral collaboration on capability evaluation and systemic-risk assessment. ## §3 International-Coordination Integration For multilateral AI governance: - UN AI Advisory Body (Final Report 2024 follow-up) — operator engagement with the proposed international scientific panel and the Global Fund for AI. - GPAI Working Groups — operator participation in GPAI's responsible AI / data governance / future of work / innovation working groups. - OECD AI Network of Experts — OECD AI Principles implementation evaluation participation. - UNESCO Recommendation 2021 implementation reporting through the operating jurisdiction's UNESCO national-commission. ## §4 EU AI Office and Member-State NCA Integration For EU-jurisdiction operators: - EU AI Office — for GPAI providers Article 56 Code of Practice signatory channel; serious incidents under Article 55(1)(c) reporting; market-surveillance coordination. - Member-State NCA (Notified Authorities) — for high-risk system providers and deployers the primary supervisory authority for AI Act compliance. - AI Board (Article 65) — Member-State coordination on cross-border issues. - Advisory Forum (Article 67) — civil-society + industry stakeholder consultation channel. - Scientific Panel of independent experts (Article 68) — frontier risk advisory. ## §5 US Sector Regulator Integration For US-jurisdiction operators: - FTC — for AI-related deception / unfairness / privacy enforcement. - EEOC — for AI in employment decisions. - DOT NHTSA — for AI in automotive safety. - FDA — for AI/ML in medical devices (the pre-determined change-control plan). - CFPB — for AI in consumer credit (cross- reference WIA-credit-scoring). - DOE — for AI in critical-infrastructure (cross- reference WIA-energy-cloud + WIA-distributed- energy). - DHS / CISA — for AI cybersecurity. ## §6 KR AI 기본법 and KAISI Integration For KR-jurisdiction operators: - KR 인공지능 발전 및 신뢰 기반 조성 등에 관한 기본법 — the operator's high-impact-AI / generative-AI / fundamental-AI obligations under the AI Basic Act. - KR AI 안전연구소 (KAISI) — pre-deployment evaluation cooperation. - KR PIPC — for personal-information-related AI obligations. - KR 산업통상자원부 — for industrial AI policy alignment. - KR MSIT — for AI R&D and digital-platform policy. ## §7 Workforce-Transition Partner Integration The workforce-transition partner integration: - Educational institutions — university and technical-college reskilling partnerships. - Public employment agencies — KR 고용노동부 + US Department of Labor + EU EURES + the jurisdiction's equivalent. - Sectoral training funds — labour-market retraining funds operating in the operator's sector. - Trade unions — collective-bargaining engagement on AI-driven workforce changes. - Industry associations — sector-wide reskilling programmes. ## §8 Responsible-Scaling Commitment Network Integration For frontier-AI providers participating in the voluntary commitment network: - Seoul AI Summit 2024 commitments — published scaling policies, frontier evaluations, system- card disclosure. - AI Action Summit Paris 2025 commitments — updated commitments published per the summit's declaration. - White House Voluntary Commitments (2023) for US-jurisdiction operators. - AISI Consortium voluntary commitments at national level. ## §9 External Audit and AIMS Certification The operator's ISMS is certified against ISO/IEC 27001:2022; the AIMS is certified against ISO/IEC 42001:2023. The certification body operates under ISO/IEC 17021-1; the conformity-assessment body for WIA-ai-survival-2026 operates under ISO/IEC 17065. EU AI Act Article 43 third-party conformity assessment is exercised for high-risk systems where required by Annex III. ## §10 Dual-Use Export-Control Integration For operators subject to dual-use controls: - US BIS (Bureau of Industry and Security) for EAR. - US DDTC (Directorate of Defense Trade Controls) for ITAR. - EU Member-State export-control authority for Reg (EU) 2021/821. - Wassenaar Arrangement national point-of-contact. - KR 산업통상자원부 + 전략물자관리원 for KR- jurisdiction. ## §11 Long-Term Archival Integration Records governed by the operator's retention horizons (EU AI Act Article 18 ten-year retention for high-risk systems; US OMB M-24-10 federal-use- case-inventory annual; KR AI 기본법 retention) are migrated to the long-term archive at the close of the active retention window. The archive preserves the AI-system inventory, the AIMS records, the safety-test reports, the incident records, the frontier-policy amendments, the supply-chain attestations, the dual-use licences, the workforce- transition programme records, and the audit-event trail. ## §12 Compute-Provider and Cloud-AI Integration For operators dependent on third-party compute / cloud-AI providers: - Compute-provider tier reporting per AI Act Article 53 + the operator's compute-budget declaration. - Per-provider cloud-AI service-level agreement including the provider's safety attestation. - Cross-cloud portability + vendor-lock-in risk mitigation. - Sovereign-cloud and data-localisation discipline per the operating jurisdiction's data-protection regime. ## §13 Civil-Society and Affected-Communities Integration For inclusive AI governance: - Stakeholder engagement under EU AI Act Article 27 FRIA + ISO/IEC 42001 Clause 4 interested- parties. - Affected-rights-holder consultation through published comment periods + targeted outreach. - Independent AI auditor coordination (Algorithmic Justice League, AI Now Institute, Distributed AI Research Institute, KAIST AI 정책연구센터, KOREA AI 윤리연구회, etc.) where the operator participates. - Right to lodge a complaint per AI Act Article 86 routed through the operator's complaint channel to the Member-State NCA. ## §14 Public-Sector AI Use-Case Inventory Integration For US-jurisdiction federal-agency operators: - US OMB M-24-10 federal AI use-case inventory annual submission. - Public publication of safety-impacting and rights-impacting AI use cases. For other jurisdictions: - EU public-sector AI Act Article 49 register entries. - KR 인공지능 행정기본법 administrative AI registry. ## §15 Conformance Implementations claiming PHASE-4 conformance maintain the AISI-network, EU AI Office, US sector regulator, KR AI 안전연구소, workforce-transition, responsible- scaling, dual-use, and compute-provider integrations, hold the ISO/IEC 42001 + 27001 certifications, and operate the long-term archival integration described above. --- **Document Information:** - **Version:** 1.0 - **Phase:** 4 — INTEGRATION - **Status:** Stable - **Standard:** WIA-ai-survival-2026 - **Last Updated:** 2026-04-29