Short abstract:

This presentation explores the socio-technical reconfiguration of care and its affective practices in the context of Emotion AI. It provides insights into the inscribed societal norms regarding emotional sensations and the (re-)configurations of gender.

Long abstract:

This presentation explores the socio-technical reconfiguration of care in the context of Emotion AI designed for mental health support. Emotion AI is an emerging field in the development and research of artificial intelligence for the detection, interpretation and simulation of (human) emotions. Such a technology seeks to introduce a novel dimension of intimacy into the field of healthcare, where emotional sensations are increasingly reconfigured by and deeply intertwined with AI devices. Against this backdrop, gender biases are inscribed into AI technologies, suggesting an uncertain shift in the socio-technical reconfiguration of affective care practices in relation to gender. Drawing on feminist STS, this presentation thus places particular emphasis on the connections between inscribed gender roles and the embedded "feeling rules" (Hochschild 1983) of care as manifested in and through AI technologies.

By examining an AI-mediated chatbot applied to simulate romantic and therapeutic relationships, I provide empirical insights into how care and its affects are reconfigured by Emotion AI. Central to this exploration is the goal of uncovering the implicit societal 'feeling rules' guiding these mediations and the underlying gender biases that extend into the realm of ‘more-than-human care’. Based on auto-ethnographic fieldwork, I discuss how AI-mediated care inadvertently perpetuates and reinforces existing heteronormative gender stereotypes.