The bot did it: Theorizing covert hate speech as a flight from agency in the digital mediasphere

New theories of user agency within algorithmic digital environments abound. Averse to humanist tradition, these theories emphasize non-human elements in mediating user agency and action opportunities online brought about by algorithmic digital technology. In this article, I foreground disturbing aspects of digital culture, such as covert racism illuminated by Microsoft chatbot Tay’s incident, to scrutinize shared assumptions originating in posthumanist and new materialist philosophies behind this trend. Revisiting Neff and Nagy’s case for understanding Tay’s incident in terms of symbiotic agency, I show how it cuts across two distinct understandings of agency: agency that is distributed among human and non-human entities on the one hand and the indivisible or inescapable human agency posited by the constitutivist theory in metaethics on the other. Focusing on communicative and cultural contexts helps to see why the attempts to go beyond the human are contradictory: blurring the boundary between human and non-human agents by conceiving agency as inherently distributed or relational leaves us blind to the apparent fact that machine agency traces back to human agency. In my turn, I propose a psychoanalytical reading of Tay as a case of interpassive disavowal – users performatively fleeing their agency. This reading moderates Kantian constitutivist theory and helps us advance in agency debates within media and communication studies by a constructive, culturally focused critique.