Of Monsters and Cisgenders: Trans Data Poisoning Against Algorithmic Surveillance

Christoffer Koch Andersen, Mallika Dharmaraj

Keywords: data poisoning, algorithmic surveillance, trans, monstrosity

TL;DR: In this paper, we introduce the concept of trans data poisoning as a collective strategy of catalysing the unintelligibility and monstrosity of trans bodies to destabilise and resist algorithmic surveillance.

Abstract: In this paper, we propose a collective strategy to resist algorithms located at the trans body that we call trans data poisoning. Drawing from Trans Studies literature, what we understand as the inherent “monstrosity” of transness may offer a novel tactic of “data poisoning” against algorithmic surveillance, as articulated by scholars in Artificial Intelligence (AI) and Machine Learning (ML) Ethics. We begin with an overview of trans scholarship on monstrosity in combination with technical AI/ML data poisoning literature. Secondly, by bridging these two areas of scholarship and discussing their overlaps, we introduce the concept of trans data poisoning as a collective strategy of catalysing the unintelligibility and monstrosity of trans bodies to destabilise and resist algorithmic surveillance — followed by two models of trans data poisoning, implicit and explicit, across India and Denmark. Finally, we discuss the collective potentials of mobilising the monstrosity of transness as a strategy to poison algorithmic systems and how future research may be informed by this counterstrategy for resisting violence to promote trans liveability and algorithmic justice.

Submission Number: 13