Read the original document by opening this link in a new tab.
Table of Contents
1. Abstract
2. Introduction
3. The Myth of Objectivity
4. Beyond Bias
5. Power in Data
6. Governing Artificial Intelligence
7. Technical Challenges and Bias
8. Conclusion
Acknowledgement
References
Summary
The paper addresses the potential of bias in Artificial Intelligence systems, highlighting the impact on societal equality and justice. It reviews efforts to ensure fairness and bias mitigation in AI systems from different domains. The complexity associated with defining policies for dealing with bias is discussed, along with technical challenges. The authors emphasize the importance of critical intersectional feminist and critical race theories in dealing with bias and fairness in AI systems. The paper concludes by stressing the need to involve those directly impacted by AI systems in the design and testing processes.