Strategische positionering
ACF vs AI Act / ISO 42001 / NIST AI RMF / AVG

ACF vervangt bestaande AI-standaarden niet.

Het maakt ze operationeel.

Elk toonaangevend referentiekader reguleert AI. ACF® werkt een laag dieper: op de agentic beslissing. Deze pagina laat zien hoe elk van de 17 ACF®-kaarten zich verhoudt tot de EU AI Act, ISO/IEC 42001, het NIST AI RMF, de AVG en COBIT — zodat elk compliance-team ACF binnen enkele minuten in zijn bestaande audit trail kan integreren.

Executive view

Hoe ACF zich verhoudt, in één alinea.

De dominante standaarden in elk governance-domein beantwoorden de vraag. ACF beantwoordt de volgende vraag — wat eraan te doen, op het beslissingsniveau.

BehoefteDominant bestaand antwoordWat ACF toevoegt
AI complianceEU AI ActAI Act + operationalisation at the decision layer
AI managementISO/IEC 42001ISO 42001 + agentic decision control
AI risk managementNIST AI RMFRisk + agentic autonomy & criticality
Data protectionGDPRGDPR + signed automated arbitrations
Gedetailleerde correspondentiematrix

Elke ACF-kaart gekoppeld aan AI Act, ISO 42001, NIST AI RMF, AVG en COBIT.

Behoudende mappings: wanneer een kaart meerdere artikelen raakt, citeren we het primaire. De volledige mapping (inclusief secundaire referenties) is opgenomen in de ACF®-toolkithandleiding.

ACF-kaartEU AI ActISO/IEC 42001NIST AI RMFAVGCOBIT 2019
ACF-00
Sovereignty Score
Measures the decisional sovereignty retained when deploying autonomous systems.
Art. 9Risk management systemClause 6.1.2AI risk assessmentMAP-3AI risks and benefitsArt. 35DPIAEDM-01Governance framework
ACF-01
Decision Map
Maps the agent's decisions and the human approval chain that governs them.
Art. 14Human oversightClause 8.4 / A.6AI system operationGOVERN-1.1Defined rolesArt. 22Automated decision-makingEDM-03Risk optimisation
ACF-02
Criticality Matrix
Classifies each agent by criticality, impact and irreversibility.
Art. 6 + Annex IIIHigh-risk classificationClause 6.1.2Risk categorisationMAP-2CategorisationArt. 35DPIA thresholdAPO-12Managed risk
ACF-03
Agentic Constitution
Internal charter defining who decides what, how and within which limits.
Art. 5 + Art. 26Prohibited practices + deployer dutiesClause 5.2AI policyGOVERN-2Cultivate cultureArt. 25Privacy by designEDM-01Governance setting
ACF-04
Agent Card
Operational identity of each agent: scope, data, tools, autonomy level.
Art. 11 + Art. 26(6)Technical documentation + logs retentionClause 7.5 + 8.1Documented information / operational planningMAP-1System contextArt. 30Records of processingBAI-09Managed assets
ACF-05
Supervision & Governance
Continuous supervision mechanisms with the DDAO role as the human pivot.
Art. 14 + Art. 26(5)Human oversight + deployer monitoringClause 5.3 + 9.1Roles + monitoringGOVERN-3 / MANAGE-2.3Workforce / ongoing monitoringArt. 22 + Art. 37-39Automated decisions + DPOMEA-02Internal control
ACF-06
Kill Switch
Emergency stop procedure for a drifting agent.
Art. 14(4) + Art. 26(5)Stop button + suspend obligationClause 8.3Operational controlsMANAGE-4DecommissioningArt. 22(3)Right to contest / withdrawDSS-02Service requests & incidents
ACF-07
First Agent Briefing
Qualification dossier before the first production deployment.
Art. 11–13 + Art. 17Documentation + QMSClause 8.1 + 6.2Operational planning + objectivesMAP-2 + GOVERN-4Categorisation + pre-deploy testingArt. 30 + Art. 35Records + DPIABAI-01Managed programmes
ACF-08
Agentic Decision Register
Cryptographic ledger of every decision the agent has taken.
Art. 12 + 19 + 26(6)Logging + retention (6 months min.)Clause 9.1 + 7.5.3Monitoring + control of recordsMEASURE-2Performance & trustworthinessArt. 30Records of processingMEA-01Performance monitoring
ACF-09
Action & Improvement Plan
Post-deployment continual improvement plan, driven by the DDAO.
Art. 9(4) + Art. 17Continuous risk mgmt + QMSClause 10.1 + 10.2Nonconformity + continual improvementMANAGE-2Risk treatmentArt. 24 + Art. 32Responsibility + securityBAI-08Managed knowledge
ACF-10
30-day Governance Audit
Periodic internal audit demonstrating operational mastery.
Art. 17 + Art. 71QMS audit + post-market monitoringClause 9.2 + 9.3Internal audit + management reviewGOVERN-5 + MANAGE-3.1Engagement + risk treatment reviewArt. 32Security auditMEA-02 + MEA-03Internal control + compliance
ACF-11
Agentic Risk Assessment
Risk analysis specific to agents: drift, hallucination, escalation.
Art. 9Risk management systemClause 6.1.2AI-specific risksMAP-3 + MAP-4Risks & benefits + trustworthinessArt. 35DPIAAPO-12Managed risk
ACF-12
Agent Mandate
Formal, enforceable delegation of decision-making power granted to the agent.
Art. 16 + 17 + 26Provider & deployer dutiesClause 5.3Roles & authoritiesGOVERN-3 + GOVERN-6Workforce + external communicationsArt. 28 + 24Processor + controller responsibilityAPO-05Managed portfolio
ACF-13
Guided Case Study
Worked case study, step by step, for training and mock audits.
Art. 6 + 13 + Annex IIISector examples + transparencyClause 7.2 + 7.3Competence + awarenessMAP-2CategorisationArt. 22Worked profiling examplesBAI-05Organisational change
ACF-14
Teacher Guide
For instructors: lesson plans, answer keys, sample exams.
Art. 4AI literacyClause 7.2 + 7.3Competence + awarenessGOVERN-1.6 + GOVERN-6Workforce literacyArt. 39DPO training dutyAPO-07Managed human resources
ACF-15
Governance Simulation
Sandbox exercise: replay a crisis to measure governance resilience.
Art. 9 + Art. 57-63Risk mgmt + regulatory sandboxesClause 9.1 + 6.2Monitoring + planned objectivesMANAGE-3 + MEASURE-3Risk treatment evaluationArt. 32Security testingBAI-06Managed IT changes
ACF-16
Responsibility by Design
Cross-cutting principle: accountability is wired in from design onwards.
Art. 5 + 13 + 16(b)Accountability + transparencyClause 5.2AI policy & accountabilityGOVERN-1 + MANAGE-1Accountability + risk managementArt. 5(2) + 24 + 25Accountability + by designEDM-01Governance setting

Bronnen: EU AI Act (Verordening 2024/1689) · ISO/IEC 42001:2023 · NIST AI RMF 1.0 (2023) · AVG (Verordening 2016/679) · COBIT 2019.

De onderscheidende boodschap

Bestaande kaders besturen AI. ACF bestuurt de beslissingen van autonome systemen.

De EU AI Act vertelt u dat het AI-systeem gereguleerd is. ISO/IEC 42001 vertelt u hoe u uw AI-portfolio beheert. Het NIST AI RMF vertelt u welke risico's u moet blootleggen. De AVG vertelt u wie toestemming moet geven.

Zodra een agent handelt — een prijs kiest, krediet verleent, een vlucht boekt, een API aanroept — vertelt geen van deze kaders u wie die specifieke beslissing heeft goedgekeurd, tegen welke constitutie, met welke kill switch, door wie geaudit en hoe lang bewaarbaar.

Die laag is wat ACF formaliseert. De 17 kaarten zijn het operationele vocabulaire ervan.

Categoriepositionering

Waarom agentic governance een eigen kader verdient.

Elk aangrenzend domein heeft al een dominant kader. De opkomst van autonome agents heeft een nieuw domein gecreëerd — en ACF is daarvan het referentiekader.

DomeinDominant referentiekader
CybersecurityISO/IEC 27001
PrivacyGDPR
Artificial IntelligenceISO/IEC 42001
Regulated AIEU AI Act
Agentic governanceACF®

Wilt u van positionering naar operaties?

ACF Compliance — het SaaS-product — implementeert elke bovenstaande mapping als een tenant-geïsoleerd cryptografisch register. Zelfevaluatie is gratis.

Open ACF ComplianceProbeer de gratis diagnose