Standards Alignment — ISO 42001, NIST AI RMF, EU AI Act, IEEE 7000

COMPEL aligns with major AI governance standards: ISO/IEC 42001:2023, NIST AI RMF 1.0, EU AI Act 2024/1689, and IEEE 7000-2021. This page provides detailed stage-by-stage mapping tables.

ISO/IEC 42001:2023 — AI Management Systems

ISO/IEC 42001:2023 is the first international management system standard for artificial intelligence. COMPEL operationalizes every clause across its six stages.

Stage-by-Stage ISO 42001 Mapping

• Calibrate: Clause 4 (Context), Clause 6 (Planning), Annex A gap analysis
• Organize: Clause 5 (Leadership), Clause 7 (Support), Annex A — A.2, A.3
• Model: Clause 6 (Planning), Annex A — A.4, A.5, A.6, A.7
• Produce: Clause 8 (Operation), Annex A — A.8, A.9, A.10
• Evaluate: Clause 9 (Performance evaluation), Clause 9.3 (Management review)
• Learn: Clause 10 (Improvement), Clause 9.1 (Monitoring and measurement)

Key ISO 42001 Requirements and COMPEL Answers

• Risk management system for AI (Annex A, A.6) — COMPEL D17 Risk Management domain with enterprise AI risk taxonomy
• AI system inventory and lifecycle tracking (Annex A, A.5) — COMPEL AI System Registry deployed in Produce stage
• Human oversight and intervention mechanisms (Annex A, A.8) — COMPEL Model stage human-AI collaboration blueprints
• Competence and awareness (Clause 7.2–7.3) — COMPEL Organize stage training programs and certification pathways
• Internal audit (Clause 9.2) — COMPEL Evaluate stage gate reviews and governance scorecard assessments

NIST AI RMF 1.0 — AI Risk Management Framework

The NIST AI RMF defines four core functions: GOVERN, MAP, MEASURE, and MANAGE. COMPEL’s 18-domain model provides the operational implementation layer.

Stage-by-Stage NIST AI RMF Mapping

• Calibrate: GOVERN, MAP — Assessment and Categorization
• Organize: GOVERN — Culture, Roles, Accountability
• Model: MAP — Risk ID, Context, Documentation
• Produce: MANAGE — Controls, Risk Treatment
• Evaluate: MEASURE — Testing, Benchmarking, Evaluation
• Learn: MANAGE — Monitoring, Continual Risk Response

Key NIST AI RMF Practices and COMPEL Answers

• GOVERN 1.1: Policies and processes for trustworthy AI — COMPEL Model stage AI Policy Framework and D18 Governance Structure
• MAP 1.5: AI risk assessment across lifecycle — COMPEL D17 Risk Management taxonomy applied at Calibrate and maintained through Learn
• MEASURE 2.5: AI risk measured and documented — COMPEL Evaluate stage governance scorecard and Gate E review
• MANAGE 1.3: Risk responses planned and documented — COMPEL risk treatment plans tracked through Continuous Improvement Register

EU AI Act 2024/1689

The EU AI Act is the world’s first comprehensive legal framework for AI. Non-compliance carries fines up to €35 million or 7% of global annual turnover.

Stage-by-Stage EU AI Act Mapping

• Calibrate: Article 9 (Risk management system), risk classification per Annex III
• Organize: Article 4 (AI literacy), Article 26 (Deployer obligations)
• Model: Article 13 (Transparency), Article 14 (Human oversight), Article 17 (Quality management)
• Produce: Article 9 (Risk management implementation), Article 12 (Record-keeping)
• Evaluate: Article 43 (Conformity assessment), Article 15 (Accuracy and robustness)
• Learn: Article 72 (Post-market monitoring), Article 73 (Incident reporting)

Key EU AI Act Articles and COMPEL Answers

• Article 9: Risk management system throughout AI lifecycle — COMPEL’s full six-stage cycle constitutes the risk management system
• Article 13: Transparency and provision of information — COMPEL Model stage system documentation and registry records
• Article 14: Human oversight measures — COMPEL human-AI collaboration blueprints designed in Model, implemented in Produce, verified in Evaluate
• Article 17: Quality management system — COMPEL’s six-stage cycle is the QMS; gates M, P, E, L are the control points
• Article 72: Post-market monitoring — COMPEL Learn stage monitoring, KPI tracking, and drift detection

IEEE 7000-2021 — Ethical Design

IEEE 7000-2021 establishes a model process for addressing ethical concerns during system design, introducing value-based engineering. COMPEL’s Model stage directly incorporates IEEE 7000 practices.

Stage-by-Stage IEEE 7000 Mapping

• Calibrate: Ethical context assessment, value landscape
• Organize: Ethics Board formation, D15 ownership
• Model: Value elicitation, ethical risk assessment, Gate M (primary alignment)
• Produce: Value-based design implementation
• Evaluate: Design verification, ethics review, bias testing (primary alignment)
• Learn: Ongoing ethical monitoring