TS EN ISO/IEC 12792

TS EN ISO/IEC 12792 is a terminology and conceptual framework standard for classifying the information elements used to assess and manage transparency in artificial intelligence systems. It helps AI stakeholders identify what information is needed, and distinguish the kinds of transparency-related information that support different purposes.

The standard treats transparency as a multi-dimensional concept rather than a single attribute. In general technical terms, that means considering how an AI system is built, what it is intended to do, what data it depends on, and what limits or assumptions shape its operation. This is useful when people need to understand and compare AI systems in a structured way.

How transparency information is organised

A key idea behind this taxonomy is that different stakeholders do not need the same level or type of information. Developers, providers, users, auditors, regulators, and affected persons may each require different transparency details. The standard provides a semantic organisation of information elements so those needs can be described more consistently.

According to the supplied scope, the document applies to any kind of organization and any application involving an AI system. In practical terms, that makes it a general-purpose vocabulary and classification resource rather than a sector-specific or system-specific method. It is especially relevant for documentation, governance, compliance discussions, and communication across roles.

• Classifies information elements related to AI system transparency.
• Describes the semantics of those information elements.
• Relates the elements to the objectives of different stakeholders.
• Applies broadly to organizations and applications involving AI systems.