(Im)possibilities in the Ethics of AI: Biometric Surveillance, Complicity, and Abolitionist Refusal in India and Beyond

In this paper, I address the question of “democratizing” AI by positing a discontent framework that unsettles the very possibilities inherent in “AI Ethics” as a project. I use the question of biometric surveillance in India as an attendant case study (with its centuries of Brahminical, colonial, and nationalist violence) but conclude with notes of relinquishment applicable broadly. To begin, I conduct a literature review of two key schools of AI Ethical thought—(1) a neoliberal fairness-centric first-wave (2) a staunchly sociotechnical second-wave of largely Black feminist criticism, which has sought to disrupt assumptions inherent in the former. Extending this critique, I then argue that the most ethical framework of AI in modern-day India seeks not to resolve, but to abolish and requires a politics of relinquishment from the dominant-caste Hindu (particularly those of us with ties to Silicon Valley’s technological violence). Drawing inspiration from scholars who have keenly applied nuggets from Black feminist, abolitionist, and queer theory to the realm of the digital, this paper seeks to join a critical wave of scholarship that abandons the Western-neoliberal project of “AI Ethics” as such in favor of a reimagined site of complicity, refusal and hopelessness, guided by epistemologies from the margins—in short, this paper endeavors to discuss the (im)possibilities inherent in reconceiving an Ethics of AI .