SIST EN ISO/IEC 12792:2026

SLOVENSKI STANDARD
01-januar-2026
Informacijska tehnologija - Umetna inteligenca (UI) - Taksonomija preglednosti
sistemov UI (ISO/IEC 12792:2025)
Information technology - Artificial intelligence (AI) - Transparency taxonomy of AI
systems (ISO/IEC 12792:2025)
Informationstechnologie - Künstliche Intelligenz - Transparenz-Taxonomie von
KI‑Systemen (ISO/IEC 12792:2025)
Technologies de l'information - Intelligence artificielle - Taxonomie pour la transparence
des systèmes d’IA (ISO/IEC 12792:2025)
Ta slovenski standard je istoveten z: EN ISO/IEC 12792:2025
ICS:
01.040.35 Informacijska tehnologija. Information technology
(Slovarji) (Vocabularies)
35.020 Informacijska tehnika in Information technology (IT) in
tehnologija na splošno general
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.

EUROPEAN STANDARD EN ISO/IEC 12792

NORME EUROPÉENNE
EUROPÄISCHE NORM
November 2025
ICS 35.020; 01.040.35
English version
Information technology - Artificial intelligence (AI) -
Transparency taxonomy of AI systems (ISO/IEC
12792:2025)
Technologies de l'information - Intelligence artificielle Informationstechnologie - Künstliche Intelligenz -
(IA) - Taxonomie pour la transparence des systèmes Transparenz-Taxonomie von KI-Systemen (ISO/IEC
d'IA (ISO/IEC 12792:2025) 12792:2025)
This European Standard was approved by CEN on 29 August 2025.

CEN and CENELEC members are bound to comply with the CEN/CENELEC Internal Regulations which stipulate the conditions for
giving this European Standard the status of a national standard without any alteration. Up-to-date lists and bibliographical
references concerning such national standards may be obtained on application to the CEN-CENELEC Management Centre or to
any CEN and CENELEC member.
This European Standard exists in three official versions (English, French, German). A version in any other language made by
translation under the responsibility of a CEN and CENELEC member into its own language and notified to the CEN-CENELEC
Management Centre has the same status as the official versions.

CEN and CENELEC members are the national standards bodies and national electrotechnical committees of Austria, Belgium,
Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy,
Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal, Republic of North Macedonia, Romania, Serbia,
Slovakia, Slovenia, Spain, Sweden, Switzerland, Türkiye and United Kingdom.

CEN-CENELEC Management Centre:
Rue de la Science 23, B-1040 Brussels
© 2025 CEN/CENELEC All rights of exploitation in any form and by any means
Ref. No. EN ISO/IEC 12792:2025 E
reserved worldwide for CEN national Members and for
CENELEC Members.
Contents Page
European foreword . 3

European foreword
This document (EN ISO/IEC 12792:2025) has been prepared by Technical Committee ISO/IEC JTC 1
"Information technology" in collaboration with Technical Committee CEN-CENELEC/ JTC 21 “Artificial
Intelligence” the secretariat of which is held by DS.
This European Standard shall be given the status of a national standard, either by publication of an
identical text or by endorsement, at the latest by May 2026, and conflicting national standards shall be
withdrawn at the latest by May 2026.
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. CEN-CENELEC shall not be held responsible for identifying any or all such patent rights.
Any feedback and questions on this document should be directed to the users’ national standards
body/national committee. A complete listing of these bodies can be found on the CEN and CENELEC
websites.
According to the CEN-CENELEC Internal Regulations, the national standards organizations of the
following countries are bound to implement this European Standard: Austria, Belgium, Bulgaria,
Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland,
Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal, Republic of
North Macedonia, Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Türkiye and the
United Kingdom.
Endorsement notice
The text of ISO/IEC 12792:2025 has been approved by CEN-CENELEC as EN ISO/IEC 12792:2025
without any modification.
International
Standard
ISO/IEC 12792
First edition
Information technology — Artificial
2025-11
intelligence (AI) — Transparency
taxonomy of AI systems
Technologies de l'information — Intelligence artificielle (IA) —
Taxonomie pour la transparence des systèmes d’IA
Reference number
ISO/IEC 12792:2025(en) © ISO/IEC 2025

ISO/IEC 12792:2025(en)
© ISO/IEC 2025
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting on
the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address below
or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: +41 22 749 01 11
Email: copyright@iso.org
Website: www.iso.org
Published in Switzerland
© ISO/IEC 2025 – All rights reserved
ii
ISO/IEC 12792:2025(en)
Contents Page
Foreword .v
Introduction .vi
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Symbols and abbreviated terms. 4
5 Overview . 4
5.1 General .4
5.2 Organization and usage of the taxonomy .4
5.3 Constraints on transparency disclosures .5
5.4 Concept of transparency .6
6 Stakeholders’ needs and transparency objectives . 7
6.1 General .7
6.2 Transparency objectives and goals .8
6.3 Selected stakeholder roles in transparency .9
7 Context-level taxonomy .11
7.1 General .11
7.2 Societal context .11
7.2.1 General .11
7.2.2 Labour practices .14
7.2.3 Consumer needs . 15
7.3 Environmental context .16
7.3.1 General .16
7.3.2 Particular environmental disclosures .17
7.3.3 Organizational disclosures .18
8 System-level taxonomy . 19
8.1 General .19
8.2 Basic information .19
8.3 Organizational processes . 20
8.3.1 General . 20
8.3.2 Governance . 20
8.3.3 Management system .21
8.3.4 Risk management .21
8.3.5 Quality management .21
8.4 Applicability .21
8.4.1 General .21
8.4.2 Intended purposes . 22
8.4.3 Capabilities . 22
8.4.4 Functional limitations . 22
8.4.5 Recommended uses . 22
8.4.6 Precluded uses . 22
8.5 Overview of technical characteristics . 22
8.5.1 General . 22
8.5.2 Expected inputs and outputs . 23
8.5.3 Production data . 23
8.5.4 Logging and storing . 23
8.5.5 System decomposition . 23
8.5.6 Application programming interface .24
8.5.7 Human factors .24
8.5.8 Deployment methods .24
8.5.9 Configuration management .24
8.6 Access to internal elements . 25

© ISO/IEC 2025 – All rights reserved
iii
ISO/IEC 12792:2025(en)
8.7 Quality and performance . 26
8.7.1 General . 26
8.7.2 Verification and validation processes . 26
8.7.3 Runtime measurements . 26
8.7.4 Comparison with alternative systems .27
9 Model-level taxonomy .27
9.1 General .27
9.2 Basic information .27
9.3 Usage and model interplay . . 28
9.3.1 Processing performed by the model . 28
9.3.2 Dependence on other models . 28
9.3.3 Coherence with AI system’s intended purposes. 28
9.4 Technical characteristics . 28
9.4.1 Type of technology used . 28
9.4.2 Features extracted from input data . 28
9.4.3 Algorithm used for processing . 28
9.4.4 Procedure for building the model . 28
9.4.5 Hyperparameters . 29
9.4.6 Input and output formats . 30
9.4.7 Compute hardware . 30
9.4.8 Computational costs . .31
9.4.9 Models in evolutive systems .31
9.5 Data used .31
9.6 Functional correctness .32
10 Dataset-level taxonomy .32
10.1 General .32
10.2 Basic information . 33
10.3 Data provenance . 33
10.4 Data properties . 35
10.5 Dataset domain and purposes . 36
10.5.1 General . 36
10.5.2 Language data domain details . 36
10.5.3 Vision data domain details .37
10.6 Data biases and limitations .37
10.7 Societal considerations .37
10.8 Data preparation performed . 38
10.9 Dataset maintenance . 39
Annex A (informative) Examples of transparency templates.40
Annex B (informative) Examples of stakeholder roles in transparency . 41
Bibliography .43

© ISO/IEC 2025 – All rights reserved
iv
ISO/IEC 12792:2025(en)
Foreword
ISO (the International Organization for Standardization) and IEC (the International Electrotechnical
Commission) form the specialized system for worldwide standardization. National bodies that are
members of ISO or IEC participate in the development of International Standards through technical
committees established by the respective organization to deal with particular fields of technical activity.
ISO and IEC technical committees collaborate in fields of mutual interest. Other international organizations,
governmental and non-governmental, in liaison with ISO and IEC, also take part in the work.
The procedures used to develop this document and those intended for its further maintenance are described
in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types
of document should be noted. This document was drafted in accordance with the editorial rules of the ISO/
IEC Directives, Part 2 (see www.iso.org/directives or www.iec.ch/members_experts/refdocs).
ISO and IEC draw attention to the possibility that the implementation of this document may involve the
use of (a) patent(s). ISO and IEC take no position concerning the evidence, validity or applicability of any
claimed patent rights in respect thereof. As of the date of publication of this document, ISO and IEC had not
received notice of (a) patent(s) which may be required to implement this document. However, implementers
are cautioned that this may not represent the latest information, which may be obtained from the patent
database available at www.iso.org/patents and https://patents.iec.ch. ISO and IEC shall not be held
responsible for identifying any or all such patent rights.
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions
related to conformity assessment, as well as information about ISO's adherence to the World Trade
Organization (WTO) principles in the Technical Barriers to Trade (TBT) see www.iso.org/iso/foreword.html.
In the IEC, see www.iec.ch/understanding-standards.
This document was prepared by Joint Technical Committee ISO/IEC JTC 1, Information technology,
Subcommittee SC 42, Artificial intelligence, in collaboration with the European Committee for Standardization
(CEN) Technical Committee CEN/CLC/JTC 21, Artificial Intelligence, in accordance with the Agreement on
technical cooperation between ISO and CEN (Vienna Agreement).
Any feedback or questions on this document should be directed to the user’s national standards
body. A complete listing of these bodies can be found at www.iso.org/members.html and
www.iec.ch/national-committees.

© ISO/IEC 2025 – All rights reserved
v
ISO/IEC 12792:2025(en)
Introduction
The objectives of this document include:
— improving trustworthiness, accountability and communication among different AI stakeholders,
including partners in a supply chain, customers, users, society and regulators, by establishing a consistent
terminology around transparency of AI systems;
— providing AI stakeholders with information about different elements of transparency with their
relevance and possible limitations to different use cases and target audience;
— serving as a basis for developing technology-specific, industry-specific or region-specific standards for
transparency of AI systems.
Transparency for AI systems is the property of a system that stakeholders receive relevant information
about the system. This can include information on items such as system features, limitations, data, system
design and design choices (see ISO/IEC 22989:2022, 5.15.8).
Transparency enables relevant stakeholders to have access to information, so they can better understand
how an AI system is developed, deployed and can be used. For example, this allows an AI customer (such as
an AI user) of an AI system to determine if it is appropriate for their situation and supports an AI auditor in
assessing if the system complies with conformity requirements.
A standardized transparency taxonomy of AI systems helps people with different backgrounds to better
understand each other by using the same terminology. This in turn supports an improved understanding
of AI systems, and provides a foundation for developing interoperable and coherent transparency related
standards.
This document is structured as follows:
— Clause 5 provides an overview of this document and describes the concept of transparency of AI systems;
— Clause 6 discusses how transparency needs can vary depending on the AI system context and on the
stakeholders involved;
— Clause 7 discusses transparency items that describe the context of the AI system;
— Clause 8 pertains to describing transparency disclosures at the AI system level;
— Clause 9 focuses on documenting the internal functioning of the system;
— Clause 10 offers guidance on the documentation of datasets as stand-alone items.
AI systems often affect society and the environment (see Clause 7). However, it is also true that society and
the environment can affect the performance of an AI system. This can include various aspects, such as:
— environmental conditions that affect the functioning and longevity of an AI system;
— introduction or reinforcement of unwanted bias;
— organizational practices that can cause poor outcomes;
— effects such as concept drift (concept drift occurs when there is a change in the functional relationship
between the model’s input and output) and data drift (data drift is change in the statistical properties
and characteristics of the input data);
— formation of unwanted feedback loops (e.g. negative reinforcement of discriminatory patterns), which
can be especially problematic in the case of continuous learning.
While these are important items for consideration, a detailed analysis is not provided in this document.
Some of these aspects are further detailed in ISO/IEC 23894, ISO/IEC 27701, and ISO/IEC 29134.

© ISO/IEC 2025 – All rights reserved
vi
International Standard ISO/IEC 12792:2025(en)
Information technology — Artificial intelligence (AI) —
Transparency taxonomy of AI systems
1 Scope
This document specifies a taxonomy of information elements to assist AI stakeholders with identifying
and addressing the needs for transparency of AI systems. The document describes the semantics of the
information elements and their relevance to the various objectives of different stakeholders.
This document is applicable to any kind of organization and application involving an AI system.
2 Normative references
The following documents are referred to in the text in such a way that some or all of their content constitutes
requirements of this document. For dated references, only the edition cited applies. For undated references,
the latest edition of the referenced document (including any amendments) applies.
ISO/IEC 22989:2022, Information technology — Artificial intelligence — Artificial intelligence concepts and
terminology
3 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO/IEC 22989 and the following apply.
ISO and IEC maintain terminology databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https:// www .iso .org/ obp
— IEC Electropedia: available at https:// www .electropedia .org/
3.1
attestation
issue of a statement, based on a decision, that fulfilment of specified requirements (3.14) has been
demonstrated
Note 1 to entry: The resulting statement is intended to convey the assurance that the specified requirements have
been fulfilled. Such an assurance does not, of itself, provide contractual or other legal guarantees.
Note 2 to entry: First-party attestation and third-party attestation are distinguished by the terms declaration,
certification and accreditation, but there is no corresponding term applicable to second-party attestation.
[SOURCE: ISO/IEC 17000:2020, 7.3; modified — removed 'referred to in this document as a "statement of
conformity"' from Note 1.]
3.2
carbon footprint
carbon footprint of a product
CFP
sum of greenhouse gas emissions and greenhouse gas removals in a product system, expressed as carbon
dioxide equivalents and based on a life cycle assessment using the single impact category of climate change
[SOURCE: ISO 14050:2020, 3.11.1; modified — added the preferred term "carbon footprint"]

© ISO/IEC 2025 – All rights reserved
ISO/IEC 12792:2025(en)
3.3
confidential data
data to which only a limited number of persons have access and which are meant for restricted use
[SOURCE: ISO 5127:2017, 3.1.10.18; modified — removed Note 1 to entry]
3.4
conformity assessment body
body that performs conformity assessment activities, excluding accreditation
[SOURCE: ISO/IEC 17000:2020, 4.6]
3.5
data breach
compromise of security that leads to the accidental or unlawful destruction, loss, alteration, unauthorized
disclosure of, or access to protected data transmitted, stored, or otherwise processed
[SOURCE: ISO/IEC 27040:2024, 3.5.2]
3.6
data protection
implementation of administrative, technical, or physical measures to guard against the unauthorized
access to data
[SOURCE: ISO/IEC 2382:2015, 2126371; modified — removed all Notes to entry.]
3.7
digital fingerprint
bit sequence generated from a digital document using an algorithm that uniquely identifies the original
document
Note 1 to entry: Any digital document modification will produce a different fingerprint.
[SOURCE: ISO 14641:2018, 3.15]
3.8
evolutive system
AI system whose behaviour can change without an explicit action from its AI developer
Note 1 to entry: Examples of evolutive systems include AI systems with a storage component, AI systems that leverage
user-side actions and feedback, and AI systems involving continuous learning.
Note 2 to entry: Change of behaviour does not imply a change of goal but can result in a better or worse fulfilment of
the targeted goal.
3.9
feature
measurable property of an object or event with respect to a set of characteristics
Note 1 to entry: Features play a role in training and prediction.
Note 2 to entry: Features provide a machine-readable way to describe the relevant objects. As the algorithm will not
go back to the objects or events themselves, feature representations are designed to contain all useful information.
[SOURCE: ISO/IEC 23053:2022, 3.3.3; modified — removed domain tag]

© ISO/IEC 2025 – All rights reserved
ISO/IEC 12792:2025(en)
3.10
life cycle impact assessment
LCIA
environmental impact assessment
phase of life cycle assessment aimed at understanding and evaluating the magnitude and significance of the
potential environmental impacts for a product system throughout the life cycle of the product
[SOURCE: ISO 14050:2020, 3.6.5; modified — added admitted term "environmental impact assessment"]
3.11
object of conformity assessment
object
entity to which specified requirements (3.14) apply
EXAMPLE Product, process, service, system, installation, project, data, design, material, claim, person, body or
organization, or any combination thereof.
Note 1 to entry: The term “body” is used in this definition to refer to conformity assessment bodies and accreditation
bodies. The term “organization” is used in its general meaning and may include bodies according to the context. The
more specific ISO/IEC Guide 2 definition of an organization as a body based on membership is not applicable to the
field of conformity assessment.
[SOURCE: ISO/IEC 17000:2020, 4.2; modified — "in this document" to "in this definition" in Note 1 to entry]
3.12
power usage effectiveness
PUE
ratio of the data centre total energy consumption to information technology equipment energy consumption,
calculated, measured or assessed across the same period
Note 1 to entry: Sometimes the inverse value of PUE, referred to as Data Centre Infrastructure Efficiency (DCiE), is used.
[SOURCE: ISO/IEC 30134-2:2016, 3.1.3]
3.13
scope of attestation
range or characteristics of objects of conformity assessment (3.11) covered by attestation (3.1)
[SOURCE: ISO/IEC 17000:2020, 7.4]
3.14
specified requirement
need or expectation that is stated
Note 1 to entry: Specified requirements can be stated in normative documents such as regulations, standards and
technical specifications.
Note 2 to entry: Specified requirements can be detailed or general.
[SOURCE: ISO/IEC 17000:2020, 5.1]
3.15
taxonomy
systematic classification of items into generic groups based on factors possibly common to several of the items
[SOURCE: ISO 14224:2016, 3.92]
3.16
water footprint
metric(s) that quantifies the potential environmental impacts related to water
[SOURCE: ISO 14050:2020, 3.10.1]

© ISO/IEC 2025 – All rights reserved
ISO/IEC 12792:2025(en)
4 Symbols and abbreviated terms
AI artificial intelligence
ML machine learning
API application programming interface
URL universal resource locator
REST representational state transfer
JSON JavaScript object notation
GUI graphical user interface
YAML YAML ain’t markup language
5 Overview
5.1 General
Clause 5 provides an overview of this document and describes the concept of transparency of AI systems, in
particular, in relation to other trustworthiness characteristics.
5.2 Organization and usage of the taxonomy
This document describes a multi-faceted taxonomy of transparency for AI systems. Clause 6 discusses how
transparency needs can vary, depending on the AI system context and the role of different AI stakeholders. The
remainder of this document covers multiple levels and viewpoints on the system. It is organized in four parts:
— Clause 7 describes transparency taxonomy elements (referred to simply as "taxonomy elements")
informing on the context of the AI system (e.g. transparency needs in the case of labour stakeholders).
— Clause 8 pertains to describing transparency disclosures at the AI system level (e.g. its recommended
and precluded uses, or API documentation).
— Clause 9 focuses on the internal functioning of the system, at the level of model components and
algorithmic processes.
— Clause 10 offers guidance on the documentation of datasets as stand-alone items, considering that the
AI system’s life cycle involves those datasets (thereby also worth transparency considerations), but also
that datasets can be reused as is across several AI systems for different use cases.
For a given AI system, all four parts of the taxonomy (Clauses 7 to 10) can be considered, as they provide
complementary information. An illustration of this document's structure can be found in Figure 1.

© ISO/IEC 2025 – All rights reserved
ISO/IEC 12792:2025(en)
Figure 1 — Overview of the document structure
Since the development of an AI system can involve different stakeholders (data provider, AI developer, AI
platform provider, etc.) and since an AI system can be integrated into a larger AI system in a modular way,
this taxonomy is also designed to be modular. In particular, the documentation of an AI system can include
references to the documentation (entire, or only specific parts) of another AI system.
For example, in the case of general-purpose models released as open-source assets, it can be useful to
document the models according to Clause 9 but also Clause 8, in order to cover recommended and precluded
uses. Then another organization, when integrating the general-purpose model as an internal component of a
more complex AI system, can produce a new documentation for that AI system, with:
— its own documentation according to Clause 8 (e.g. recommended and precluded uses of the whole AI
system);
— a reference to the general-purpose model’s documentation according to Clause 9;
— a new section, documented according to Clause 9, for any custom model developed for that AI system;
— but discarding the Clause 8 information on the general-purpose model if deemed irrelevant in the context
of that particular AI system. An example of irrelevant information in this context is the recommended
or precluded uses of the general-purpose model, when it includes considerations pertaining to alternate
uses that are beyond the one use that is made of the general-purpose model within that AI system.
5.3 Constraints on transparency disclosures
Taxonomy elements described in all four parts of this document (Clauses 7 to 10) are elements that are useful
to consider in at least some cases, but do not necessarily apply to all use cases and systems. In particular:
— There can be taxonomy elements for which disclosure is not appropriate in a given use case. For
example, releasing the source code is not appropriate if it includes proprietary code (8.6). Indication
of who funded a dataset's creation can be impossible if the dataset has been created by volunteers or if
disclosure is prevented by contractual obligations (10.3). Similarly, revealing how much data labelers
were remunerated can run counter to privacy concerns (10.8).
— There can be taxonomy elements that only exist for a given type of system. For example, information on
the language variety is not necessarily applicable to a computer vision system (10.5).

© ISO/IEC 2025 – All rights reserved
ISO/IEC 12792:2025(en)
— There can be taxonomy elements that describe processes that are not possible or not appropriate in
a given use case. For example, comparative evaluation with a similar system can be impossible when
creating a new type of AI system that has no existing alternative (8.7.4).
— Similarly, considerations for diversity and inclusion are not necessarily appropriate when producing a
dataset to be representative of one specific part of the population, such as elderly females with a speech
disorder (10.6).
— To determine the usefulness of transparency elements, an organization can also consider the costs and
benefits associated with each of the relevant transparency elements, in order to manage potential trade-
offs and define the appropriate level of transparency needed for its specific use case. For example, an
organization can consider whether transparency elements can in some cases cause more harm than
benefits to individuals and societies, and whether disclosures can be in breach of applicable laws and
regulations or contractual requirements.
5.4 Concept of transparency
Transparency is a broad concept, but within this document, the focus lies primarily on transparency
as the property of a system and, to a lesser degree, on transparency as the property of an organization.
Transparency for an AI system means that appropriate information about the system is made available
to relevant stakeholders (ISO/IEC 22989:2022, 5.15.8). In contrast, the transparency of an organization
(referred to as organizational transparency in this document) is related to how appropriate activities and
decisions are communicated to relevant stakeholders (ISO/IEC TR 24028:2020, 10.3.1).
Organizational transparency is relevant in so far as organizational principles and processes affect an AI
system at various stages in its life cycle as describe
...