[AI’s White Supremacy: From Eugenics to Epistemological Violence

Artificial intelligence is often framed as a neutral tool plagued by algorithmic bias—a problem supposedly solvable through technical fixes such as debiasing datasets or diversifying teams. This talk argues that such diagnoses are fundamentally inadequate. To grasp the racial politics of AI, we must interrogate the epistemological structures that have historically classified, hierarchized, and governed human beings. The lecture traces continuities between eugenics, colonialism, and contemporary AI, arguing that machine learning can reactivate a eugenic logic: the belief that human traits can be quantified, classified, ranked, and optimized. AI does not simply reproduce eugenics; it can inherit its epistemological violence, transforming historically produced racial hierarchies into apparently neutral systems of knowledge and automated decision-making. Drawing on Frantz Fanon, Patricia Hill Collins, and María Lugones, the talk examines how racialization, gendering, and the production of knowledge are mutually constitutive. Fanon’s analysis of the racialized subject exposes the violence of defining Black humanity from within a white colonial order. These perspectives shift the question from “Is AI biased?” to a more fundamental one: Who has the power to define the categories through which AI knows human beings? The conference concludes by developing epistemological violence as a framework for understanding how AI can erase, distort, subordinate, or universalize marginalized knowledges—and asks what it would mean to decolonize AI at the level of knowledge itself: who produces knowledge, whose categories become universal, and who has the power to define the human in the era of AI.

In den KalenderBearbeiten