ISO/IEC TR 24368:2022

Information technology — Artificial intelligence — Overview of ethical and societal concerns

This document provides a high-level overview of AI ethical and societal concerns. In addition, this document: — provides information in relation to principles, processes and methods in this area; — is intended for technologists, regulators, interest groups, and society at large; — is not intended to advocate for any specific set of values (value systems). This document includes an overview of International Standards that address issues arising from AI ethical and societal concerns.

Titre manque

General Information

• Status: Published
• Publication Date: 18-Aug-2022
• ICS: 35.020 - Information technology (IT) in general
• Technical Committee: ISO/IEC JTC 1/SC 42 - Artificial intelligence
• Drafting Committee: ISO/IEC JTC 1/SC 42/WG 3 - Trustworthiness
• Current Stage: 6060 - International Standard published
• Start Date: 19-Aug-2022
• Completion Date: 19-Aug-2022

Standards Content (Sample)

TECHNICAL ISO/IEC TR
REPORT 24368
First edition
2022-08
Information technology — Artificial
intelligence — Overview of ethical and
societal concerns
Reference number
ISO/IEC TR 24368:2022(E)
© ISO/IEC 2022

---------------------- Page: 1 ----------------------
ISO/IEC TR 24368:2022(E)
COPYRIGHT PROTECTED DOCUMENT
© ISO/IEC 2022
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: [phone]
Email: [e-mail]
Website: www.iso.org
Published in Switzerland
ii
  © ISO/IEC 2022 – All rights reserved

---------------------- Page: 2 ----------------------
ISO/IEC TR 24368:2022(E)
Contents Page
Foreword .v
Introduction . vi
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Overview . 3
4.1 General . 3
4.2 Fundamental sources . 4
4.3 Ethical frameworks . 6
4.3.1 General . 6
4.3.2 Virtue ethics . 6
4.3.3 Utilitarianism . 6
4.3.4 Deontology . 6
5 Human rights practices . 7
5.1 General . 7
6 Themes and principles . 8
6.1 General . 8
6.2 Description of key themes and associated principles . 8
6.2.1 Accountability . 8
6.2.2 Fairness and non-discrimination . 9
6.2.3 Transparency and explainability . 9
6.2.4 Professional responsibility . 10
6.2.5 Promotion of human values. 10
6.2.6 Privacy . 11
6.2.7 Safety and security . 11
6.2.8 Human control of technology .12
6.2.9 Community involvement and development .12
6.2.10 Human centred design .13
6.2.11 Respect for the rule of law . 13
6.2.12 Respect for international norms of behaviour .13
6.2.13 Environmental sustainability . 14
6.2.14 Labour practices . 14
7 Examples of practices for building and using ethical and socially acceptable AI .15
7.1 Aligning internal process to AI principles . 15
7.1.1 General .15
7.1.2 Defining ethical AI principles the organization can adopt .15
7.1.3 Defining applications the organization cannot pursue .15
7.1.4 Review process for new projects . 15
7.1.5 Training in AI ethics . 16
7.2 Considerations for ethical review framework . 16
7.2.1 Identify an ethical issue . 16
7.2.2 Get the facts . 16
7.2.3 List and evaluate alternative actions . 17
7.2.4 Make a decision and act on it . 17
7.2.5 Act and reflect on the outcome . 17
8 Considerations for building and using ethical and socially acceptable AI .17
8.1 General . 17
8.2 Non-exhaustive list of ethical and societal considerations . 17
8.2.1 General . 17
8.2.2 International human rights . 18
8.2.3 Accountability . 18
iii
© ISO/IEC 2022 – All rights reserved

---------------------- Page: 3 ----------------------
ISO/IEC TR 24368:2022(E)
8.2.4 Fairness and non-discrimination . 18
8.2.5 Transparency and explainability . 18
8.2.6 Professional responsibility . 19
8.2.7 Promotion of human values. 20
8.2.8 Privacy . 20
8.2.9 Safety and security . 20
8.2.10 Human control of technology . 21
8.2.11 Community involvement and development . 21
8.2.12 Human centred design . 21
8.2.13 Respect for the rule of law . 21
8.2.14 Respect for international norms of behaviour .22
8.2.15 Environmental sustainability . 22
8.2.16 Labour practices .22
Annex A (informative) AI principles documents .23
Annex B (informative) Use case studies .33
Bibliography .42
iv
  © ISO/IEC 2022 – All rights reserved

---------------------- Page: 4 ----------------------
ISO/IEC TR 24368:2022(E)
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).
Attention is drawn to the possibility that some of the elements of this document may be the subject
of patent rights. ISO and IEC shall not be held responsible for identifying any or all such patent
rights. Details of any patent rights identified during the development of the document will be in the
Introduction and/or on the ISO list of patent declarations received (see www.iso.org/patents) or the IEC
list of patent declarations received (see https://patents.iec.ch).
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.
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.
v
© ISO/IEC 2022 – All rights reserved

---------------------- Page: 5 ----------------------
ISO/IEC TR 24368:2022(E)
Introduction
Artificial intelligence (AI) has the potential to revolutionise the world and carry a plethora of
benefits for societies, organizations and individuals. However, AI can introduce substantial risks and
uncertainties. Professionals, researchers, regulators and individuals need to be aware of the ethical
and societal concerns associated with AI systems and applications.
Potential ethical concerns in AI are wide ranging. Examples of ethical and societal concerns in AI
include privacy and security breaches to discriminatory outcomes and impact on human autonomy.
Sources of ethical and societal concerns include but are not limited to:
— unauthorized means or measures of collection, processing or disclosing personal data;
— the procurement and use of biased, inaccurate or otherwise non-representative training data;
— opaque machine learning (ML) decision-making or insufficient documentation, commonly referred
to as lack of explainability;
— lack of traceability;
— insufficient understanding of the social impacts of technology post-deployment.
AI can operate unfairly particularly when trained on biased or inappropriate data or where the model
or algorithm is not fit-for-purpose. The values embedded in algorithms, as well as the choice of problems
AI systems and applications are used for to address, can be intentionally or inadvertently shaped by
developers’ and stakeholders’ own worldviews and cognitive bias.
Future development of AI can expand existing systems and applications to grow into new fields and
increase the level of automation which these systems have. Addressing ethical and societal concerns
has not kept pace with the rapid evolution of AI. Consequently, AI designers, developers, deployers
and users can benefit from flexible input on ethical frameworks, AI principles, tools and methods for
risk mitigation, evaluation of ethical factors, best practices for testing, impact assessment and ethics
reviews. This can be addressed through an inclusive, interdisciplinary, diverse and cross-sectoral
approach, including all AI stakeholders, aided by International Standards that address issues arising
from AI ethical and societal concerns, including work by Joint Technical Committee ISO/IEC JTC 1, SC
42.
vi
  © ISO/IEC 2022 – All rights reserved

---------------------- Page: 6 ----------------------
TECHNICAL REPORT ISO/IEC TR 24368:2022(E)
Information technology — Artificial intelligence —
Overview of ethical and societal concerns
1 Scope
This document provides a high-level overview of AI ethical and societal concerns.
In addition, this document:
— provides information in relation to principles, processes and methods in this area;
— is intended for technologists, regulators, interest groups, and society at large;
— is not intended to advocate for any specific set of values (value systems).
This document includes an overview of International Standards that address issues arising from AI
ethical and societal concerns.
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, 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
agency
ability to define one's goals and act upon them
[SOURCE: ISO/TR 21276:2018, 3.6.2]
3.2
bias
systematic difference in treatment (3.13) of certain objects, people, or groups in comparison to others
[SOURCE: ISO/IEC TR 24027:2021, 3.2.2, modified — Removed Note to entry.]
3.3
data management
process of keeping track of all data and/or information related to the creation, production, distribution,
storage, disposal and use of e-media, and associated processes
[SOURCE: ISO 20294:2018, 3.5.4, modified — Added "disposal" to definition.]
1
© ISO/IEC 2022 – All rights reserved

---------------------- Page: 7 ----------------------
ISO/IEC TR 24368:2022(E)
3.4
data protection
legal, administrative, technical or physical measures taken to avoid unauthorized access to and use of
data
[SOURCE: ISO 5127:2017, 3.13.5.01, modified — Removed Note to entry.]
3.5
equality
state of being equal, especially in status, rights or opportunities
[SOURCE: ISO 30415:2021, 3.9, modified — Removed "outcome" from definition.]
3.6
equity
practice of eliminating avoidable or remediable differences among groups of people, whether those
groups are defined socially, economically, demographically or geographically
3.7
fairness
treatment (3.13), behaviour or outcomes that respect established facts, societal norms and beliefs and
are not determined or affected by favouritism or unjust discrimination
Note 1 to entry: Considerations of fairness are highly contextual and vary across cultures, generations,
geographies and political opinions.
Note 2 to entry: Fairness is not the same as the lack of bias (3.2). Bias does not always result in unfairness and
unfairness can be caused by factors other than bias.
3.8
cognitive bias
human cognitive bias
bias (3.2) that occurs when humans are processing and interpreting information
Note 1 to entry: Human cognitive bias influences judgement and decision-making.
[SOURCE: ISO/IEC TR 24027:2021, 3.2.4, modified — Added "cognitive bias" as preferred term.]
3.9
life cycle
evolution of a system, product, service, project or other human-made entity from conception through
retirement
[SOURCE: ISO/IEC/IEEE 12207:2017, 3.1.26]
3.10
organization
company, corporation, firm, enterprise, authority or institution, person or persons or part or
combination thereof, whether incorporated or not, public or private, that has its own functions and
administration
[SOURCE: ISO 30000:2009, 3:10]
3.11
privacy
rights of an entity (normally an individual or an organization), acting on its own behalf, to determine
the degree to which the confidentiality of their information is maintained
[SOURCE: ISO/IEC 24775-2:2021, 3.1.46]
2
  © ISO/IEC 2022 – All rights reserved

---------------------- Page: 8 ----------------------
ISO/IEC TR 24368:2022(E)
3.12
responsibility
obligation to act or take decisions to achieve required outcomes
Note 1 to entry: A decision can be taken not to act.
[SOURCE: ISO/IEC 38500:2015, 2.22, modified — Changed “and” to “or” and added Note to entry.]
3.13
treatment
kind of action, such as perception, observation, representation, prediction or decision
[SOURCE: ISO/IEC TR 24027:2021, 3.2.2, modified — Changed Note to entry to term and definition.]
3.14
safety
expectation that a system does not, under defined conditions, lead to a state in which human life, health,
property, or the environment is endangered
[SOURCE: ISO/IEC/IEEE 12207:2017, 3.1.48]
3.15
security
aspects related to defining, achieving, and maintaining confidentiality, integrity, availability,
accountability, authenticity, and reliability
Note 1 to entry: A product, system, or service is considered to be secure to the extent that its users can rely that
it functions (or will function) in the intended way. This is usually considered in the context of an assessment of
actual or perceived threats.
[SOURCE: ISO/IEC 15444-8:2007, 3.25]
3.16
sustainability
state of the global system, including environmental, social and economic aspects, in which the needs
of the present are met without compromising the ability of future generations to meet their own needs
[SOURCE: ISO/Guide 82:2019, 3.1, modified — Removed Notes to entry.]
3.17
traceability
ability to identify or recover the history, provenance, application, use and location of an item or its
characteristics
3.18
value chain
range of activities or parties that create or receive value in the form of products or services
[SOURCE: ISO 22948:2020, 3.2.11]
4 Overview
4.1 General
Ethical and societal concerns are a factor when developing and using AI systems and applications.
Taking context, scope and risks into consideration can mitigate undesirable ethical and societal
outcomes and harms. Examples of areas where there is an increasing risk for undesirable ethical and
[24]
societal outcomes and harms include the following :
— financial harm;
3
© ISO/IEC 2022 – All rights reserved

---------------------- Page: 9 ----------------------
ISO/IEC TR 24368:2022(E)
— psychological harm;
— harm to physical health or safety;
— intangible property (for example, IP theft, damage to a company’s reputation);
— social or political systems (for example, election interference, loss of trust in authorities);
— civil liberties (for example, unjustified imprisonment or other punishment, censorship, privacy
breaches).
In the absence of such considerations, there is a risk that the technology itself can levy significant social
or other consequences, with possible unintended or avoidable costs, even if it performs flawlessly from
a technical perspective.
4.2 Fundamental sources
Various sources address ethical and societal concerns specifically or in a general way. Some of these
sources are identified.
Firstly, ISO Guide 82 provides guidance to standards developers in considering sustainability in their
activities with specific reference to the social responsibility guidance of ISO 26000. This document
therefore describes social responsibility in a form that can inform activities related to standardising
trustworthy AI.
ISO 26000 provides organizations with guidance concerning social responsibility. It is based on the
fundamental practices of:
— recognizing social responsibility within an organization;
— undertaking stakeholder identification and engagement.
Without data, the development and use of AI cannot be possible. Therefore, the importance of data and
data quality makes traceability and data management a pivotal consideration in the use and development
of AI. The following data-oriented elements are at the core of creating ethical and sustainable AI:
— data collection (including the means or measures of such data collection);
— data preparation;
— monitoring of traceability;
— access and sharing control (authentication);
— data protection;
— storage control (adding, change, removal);
— data quality.
These elements impact explainability, transparency, security and privacy, especially in cases of personal
identifiable information being generated, controlled or processed. Traceability and data management
are essential considerations for an organization using or developing AI systems and applications.
ISO/IEC 38505-1 considers data value, risks and constraints in governing how data are collected, stored,
distributed, disposed of, reported on and used in organizational decision-making and procedures. The
results of data mining or machine learning activities in reporting and decision-making are regarded as
another form of data, which are therefore subject to the same data governance guidelines.
Furthermore, the description of ethical and societal concerns relative to AI systems and applications
can be based on various AI-related International Standards.
4
  © ISO/IEC 2022 – All rights reserved

---------------------- Page: 10 ----------------------
ISO/IEC TR 24368:2022(E)
ISO/IEC 22989 provides standardized AI terminology and concepts and describes a life cycle for AI
systems.
ISO/IEC 22989 also defines a set of stakeholders involved in the development and use of an AI system.
ISO/IEC 22989 describes the different AI stakeholders in the AI system value chain that include AI
provider, AI producer, AI customer, AI partner and AI subject. ISO/IEC 22989 also describes various
sub-roles of these types of stakeholders. In this document we refer to all of these different stakeholder
types collectively as stakeholders.
ISO/IEC 22989 includes “relevant regulatory and policy making authorities” as a sub-role of AI subject.
Regulatory roles for AI are currently not yet widely defined, but a range of proposals has been made
including organizations appointed by individual stakeholders; industry-representative bodies; self-
appointed civic-society actors; or institutions established through national legislation or international
treaty.
All of these features of ISO/IEC 22989 assist in the description of AI-specific ethical and societal
concerns.
As AI has the potential to impact a wide range of societal stakeholders, including future generations
impacted by changes to the environment (indirectly affected stakeholders). For example, images of
pedestrians on a sidewalk can be captured by autonomous vehicle technology, or innocent persons can
be subject to police surveillance equipment designed to survey suspected criminals.
ISO/IEC 23894 provides guidelines on managing AI-related risks faced by organizations during the
development and application of AI techniques and systems. It follows the structure of ISO 31000:2018
and provides guidance that arises from the development and use of AI systems. The risk management
system described in ISO/IEC 23894 assists in the description of ethical and societal concerns in this
document.
ISO/IEC TR 24027 describes the types and forms of bias in AI systems and how they can be measured
and mitigated. ISO/IEC TR 24027 also describes the concept of fairness in AI systems. Bias and fairness
are important for the description of AI-specific ethical and societal concerns.
ISO/IEC TR 24028 provides an introduction to AI system transparency and explainability, which are
important aspects of trustworthiness and which can impact ethical and societal concerns.
ISO/IEC TR 24030 describes a collection of 124 use cases of AI applications in 24 different application
domains. The
...