.20, and ISO/IEC 38507, 4.
Version 1. This will also help ensure that technologies will implemented with minimal bias, see ISO/IEC 38507,
4.g.4
iv
Safety ensures that “where necessary, the organization should put place appropriate systems for the ongoing management safety as
well considering how the use can reduce the exposure humans dangerous activities”, see ISO/IEC 38507, 6. Transparency “can involve communicating appropriate information about the system stakeholders (e. Transparency can also include informing stakeholders about the processing and level automation used make related
decisions”, see ISO 16759:2013, 3. For any personal data protection, intellectual property, or
licensing concerns, please contact the IEC Data Protection Officer dataprotection@iec.34
iii
Bias the “systematic difference treatment certain objects, people groups comparison others”, see ISO/IEC 22989, 3.
Additionally, transparency system can involve informing stakeholders about the details data used (e.3
v
Transparency the “open, comprehensive, accessible, clear and understandable presentation information”; “transparency and explainability
of systems (including insight into the objectives, assumptions and rules included them)” give rise important implications when systems
are used for “addressing tasks and problems that were previously performed humans […] together with adequate processes modify and
update those algorithms”.1 October 2024
i
Accountability ensures that decisions made throughout the organization, including those that are made through the use AI, are founded in
adequate governance and controls. Please note that the rights you may have under individual corporate license may not flow IEC and therefore should be
avoided.5.2. See also ISO/IEC 22989.
vi Non-licensed this document shall mean without license held IEC Secretariat; licensed shall mean with license held the IEC
Secretariat.ch or
IEC legal team through your Technical Officer.4, modified, ISO/IEC 5723:2022, 3.7.2, and ISO/IEC 42001., goals,
known limitations, definitions, design choices, assumptions, features, models, algorithms, training methods and quality assurance processes). Further reading
Stanford Index Report
Act
Executive Order (plus the White House Blueprint for Bill Rights)
Cyberspace Administration China (CAC) Draft Measures Generative Services
OECD Principles
NIST Risk Management Framework
Council Europe’s Committee Artificial Intelligence
UNESCO Recommendation the Ethics AI
World Economic Forum Governance Alliance
United Nations Global Digital Compact
ITU for Good Summits
For questions, please consult your National Committee the IEC Helpdesk for help using
IEC tools, including tools.8.3 and ISO/IEC 22989, 3.
This document will continually updated.Guidance the use artificial intelligence (AI) for IEC Committees October 2024
IEC International Electrotechnical Commission 4/4
6., what, where, when, why data
is collected and how used) produce the system and the protection personal data along with the purpose the system and how was
built and deployed.5.
ii
ISO/IEC 26927:2011, 3.2.g
