The principle ofFairness and Non-Discriminationis one of the core ethical and governance pillars emphasized inISO/IEC 42001:2023, especially inClause 4.2 (Understanding the needs and expectations of interested parties)andClause 6.1 (Actions to address risks and opportunities)where ethical risks related to AI systems are assessed.
According to the standard and accompanying PECB training guide, fairness is defined as ensuring thatAI systems do not create or perpetuate bias, and that individuals and groups aretreated equitably, with measures in place to detect and mitigate discrimination. This is often embedded in risk assessments and operational controls when managing AI systems.
In ISO/IEC 42001, organizations are required to considerethical, legal, and societal impactsof their AI systems, andFairness and Non-Discriminationis a key tenet oftrustworthy AI.
Reference: ISO/IEC 42001:2023, Clauses 4.2, 6.1.2, and 8.2.3
PECB Lead Auditor Course Guide, Section on "Ethical Principles of AI," Subsection: Fairness and Non- Discrimination