Overview
EN ISO/IEC 12792:2025 - Transparency taxonomy of AI systems defines a structured taxonomy of information elements to help AI stakeholders identify and meet transparency needs. The standard describes the semantics of transparency-related elements and explains their relevance to different stakeholder objectives. It is applicable to any organization or application involving an AI system and is published as EN ISO/IEC 12792:2025 (identical to ISO/IEC 12792:2025).
Key SEO phrases: EN ISO/IEC 12792:2025, transparency taxonomy, artificial intelligence transparency, AI systems.
Key Topics
The standard organizes transparency information at multiple levels and addresses stakeholder goals and constraints. Major technical topics and requirements include:
- Stakeholders’ needs & objectives
- Roles, transparency goals and use-case driven information needs.
- Concepts and constraints
- Definitions, semantics, and limits on transparency disclosures.
- Context-level taxonomy
- Societal, environmental and organizational context; labour and consumer-related disclosures.
- System-level taxonomy
- Basic system information, governance, management systems, risk and quality management.
- Applicability: intended purposes, capabilities, limitations, recommended/precluded uses.
- Technical characteristics: inputs/outputs, production data, logging/storage, APIs, human factors, deployment, configuration.
- Access to internal elements and runtime/validation metrics.
- Model-level taxonomy
- Model metadata, processing characteristics, dependencies, technology type and extracted features.
- Verification & validation
- Quality, performance measures, runtime measurements and comparison to alternatives.
These topics are presented as a taxonomy - a consistent vocabulary and semantic structure for transparency disclosures rather than prescriptive implementation steps.
Applications
EN ISO/IEC 12792:2025 is practical for a wide range of uses:
- AI developers & data scientists - structure model and system documentation to support explainability and traceability.
- Product managers & architects - define intended uses, limitations and integration-level transparency.
- Risk, compliance & governance teams - align disclosures with organizational risk management, quality systems and regulatory reporting.
- Auditors & assessors - use the taxonomy for transparency audits, impact assessments and conformity checks.
- Procurement & legal teams - specify transparency requirements in contracts and vendor assessments.
- Policymakers & regulators - reference a standard taxonomy when drafting rules or guidance for AI transparency.
Practical outputs include transparency reports, documentation templates, audit checklists, procurement clauses, and governance artifacts.
Related standards
This taxonomy complements other international and national AI governance, risk management and technical standards. EN ISO/IEC 12792:2025 provides a neutral vocabulary and structure that can be integrated with organizational management systems and regulatory compliance frameworks to improve consistent, auditable AI transparency.