Rethinking Artificial Intelligence: Algorithmic Bias and Ethical Issues| Mapping Scholarship on Algorithmic Bias: Conceptualization, Empirical Results, and Ethical Concerns

Abstract

As artificial intelligence (AI) becomes more seamlessly integrated into our social life, the unfair outcomes and ethical issues associated with AI and its subtechnologies have been widely discussed in scholarly work across disciplines in recent years. This study provides an overview of the conceptualization, empirical scholarship, and ethical concerns related to algorithmic bias across diverse disciplines. In doing so, the study relies on the framework of AI-mediated communication and human-AI communication, as well as topic modeling and semantic network analysis to examine the conceptualization and major thematic areas of AI bias literature. The study reveals the complexity of the concept of algorithmic bias, which extends beyond the algorithm itself. Empirical scholarship on AI and algorithmic bias revolves around conceptualizations, human perceptions, algorithm optimization, practical applications, and ethics and policy implications. Understanding and addressing the ethical challenges require a multilevel examination from the perspectives of different stakeholders. Theoretical and practical implications are further discussed in the context of AI and algorithmic justice.