AI evaluation notes from the field.

Official InstitutionEuropean CommissionAI Governance

The EU AI Act is the central official reference for risk-based AI governance in Europe. It helps enterprise teams identify obligations by AI system type, define control ownership, prepare evidence for approval, and plan monitoring before production. Use this source when mapping POC findings to compliance, procurement, and board-level rollout decisions.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Official InstitutionEuropean AI OfficeAI Governance

The European AI Office is the official coordination point for AI Act implementation and general-purpose AI oversight. It is useful for teams assessing provider obligations, model documentation, market-facing claims, and governance expectations. Use this source to connect technical review evidence with EU-level supervisory direction and implementation planning.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Official InstitutionNISTAI Risk Management

NIST AI RMF is a practical public framework for identifying, measuring, managing, and governing AI risks. It gives executives and technical teams a shared language for risk review, control design, and accountability. Use this source when turning AI pilot results into risk registers, approval evidence, and operating controls.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Official InstitutionNISTGenerative AI Risk

NIST's generative AI profile applies AI RMF thinking to risks such as harmful outputs, data exposure, provenance, misuse, and evaluation uncertainty. It is useful for LLM and generative AI pilots where normal model metrics are not enough. Use this reference to define monitoring, limits, and review evidence before deployment.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Standards BodyISOAI Governance

ISO/IEC 42001 is the key management-system standard for accountable AI governance. It helps organizations formalize policies, roles, objectives, controls, monitoring, and continual improvement. Use this source when aligning AI approval workflows with audit-ready evidence, ownership records, and management accountability across the AI lifecycle.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Standards BodyISOAI Risk Management

ISO/IEC 23894 provides international guidance for AI risk management. It is useful for structuring risk identification, treatment, documentation, and review practices. Use this source when building a risk register that connects model findings, operational exposure, control measures, and residual risk before production approval.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

International OrganisationOECDAI Governance

The OECD AI Principles remain a major international reference for trustworthy, human-centred AI. They help leaders compare internal governance policies with cross-country policy expectations and responsible AI norms. Use this source when building board-level AI principles, supplier review criteria, and policy-aligned adoption frameworks.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

International OrganisationCouncil of EuropeAI Governance

The Council of Europe's AI convention provides public-interest governance context around human rights, democracy, and rule-of-law safeguards. It is especially relevant for sensitive, public-sector, or rights-impacting AI systems. Use this source when assessing whether deployment controls address social, legal, and institutional consequences.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Supervisory BodyEuropean Data Protection BoardData Protection

The EDPB is a key European supervisory reference for data protection issues in AI systems. Its materials help teams connect privacy, transparency, fairness, lawful basis, and rights management with model evaluation. Use this source when AI pilots involve personal data or automated decision support.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Supervisory BodyUK ICOData Protection

The UK ICO's AI guidance translates data protection duties into practical questions around fairness, transparency, accountability, and explainability. It is useful for operational teams that need review evidence, user-impact analysis, and privacy controls before deployment. Use this source when evaluating AI systems that process personal data.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Supervisory BodyBank of EnglandFinancial Supervision

The Bank of England's AI materials provide financial-sector context for model risk, operational resilience, and supervisory expectations. They are useful for regulated firms that need to connect AI pilots with governance, risk ownership, and board-level oversight. Use this source when AI affects financial decisions or customer outcomes.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

International OrganisationBank for International SettlementsFinancial Supervision

BIS materials help financial institutions understand AI beyond single-model performance, including operational exposure, supervision, infrastructure risk, and systemic implications. Use this source when AI governance must address third-party dependency, market impact, resilience, and enterprise-level risk appetite.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Supervisory BodyHong Kong PCPDData Protection

Hong Kong PCPD guidance is relevant for organizations deploying AI in Hong Kong or processing personal data linked to Hong Kong users. It helps teams frame transparency, accountability, data protection, and automated decision concerns. Use this source when local compliance expectations must be reflected in AI review evidence.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

Supervisory BodyMonetary Authority of SingaporeFinancial Supervision

MAS FEAT principles provide a financial-sector reference for fairness, ethics, accountability, and transparency in AI and data analytics. They are useful when regulated firms need to translate responsible AI values into review criteria, control ownership, and approval evidence before scaling AI systems.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.

International OrganisationWorld Economic ForumAI Governance

World Economic Forum AI materials provide international business and policy context for responsible adoption, institutional readiness, and governance maturity. Use this source when executives need to compare internal AI controls with global discussion around trust, risk, competitiveness, and responsible deployment.

Disclaimer: This entry is a factual reference summary for AI governance mapping. HephaKnot does not claim endorsement by the source institution. Readers should review the original source before using it for legal, compliance, or procurement decisions.