Read the original document by opening this link in a new tab.
Table of Contents
1. Introduction
2. Scope
3. Case Study: The Impact of Familiarity Bias on Fairness
4. Fairness in AI vs. Manual Systems
5. Positive vs. Neutral Approaches Toward Fairness
Summary
This paper summarizes and evaluates various approaches, methods, and techniques for pursuing fairness in artificial intelligence (AI) systems. It examines the merits and shortcomings of these measures and proposes practical guidelines for defining, measuring, and preventing bias in AI. The paper cautions against simplistic methods for evaluating bias and offers more sophisticated alternatives. It addresses controversies and provides a common language among stakeholders. The paper describes trade-offs involving AI fairness, offers techniques for evaluating costs and benefits, and defines the role of human judgment. Discussions and guidelines are provided for AI practitioners, organization leaders, and policymakers. Real-world examples clarify concepts, challenges, and recommendations. The document covers a hypothetical case study, scope of AI fairness, familiarity bias, fairness in AI vs. manual systems, and positive vs. neutral approaches toward fairness.