Black feminism and Artificial Intelligence: the possibilities and limitations of contesting discriminatory AI from a critical social theory perspective

Abstract

Artificial Intelligence (AI) has been challenged in recent years by critical scholars inspired by American Black feminism and questions around diversity, inclusion, and justice in and through AI systems. This paper takes a closer look at the application of Black feminism as a critical social theory, which originated as a theory protesting the oppression of Black women as a group in the United States. The paper reveals the benefits and limitations of the theory in contesting AI-based sorting, classification, personalization, recommendation, and population-based predictions of different groups of technology stakeholders. Through conceptual analysis and reference to recent use cases of AI applications, the paper showcases the promise of an intersectional-type analysis and a structural perspective enabling the in-depth analysis of technology stakeholders' experiences with AI. The paper also highlights limitations of the theory in contesting AI, which inspires a discussion of constraints on critical scholarship on AI more broadly. The papers' findings and arguments are relevant to those engaging in critical conceptual, qualitative research on the design and implications of AI-based computer systems.