Paper Number

ECIS2026-1184

Paper Type

SP

Abstract

Employee perceptions of microaggressions in AI-mediated workplace interactions, including those involving avatars, remain underexplored. This study examines how individuals perceive and respond to microaggressions perpetrated by humans versus avatars in hybrid work environments. We propose a controlled laboratory experiment in which participants interact with human and avatar aggressors of different genders, while collecting multimodal behavioral data (facial expressions, gaze, posture) and self-reported emotional and cognitive responses. Contrary to expectations, no significant differences emerged between human- and avatar-delivered microaggressions regarding emotional or physiological responses. Participants reacted similarly across conditions, and the gender of the aggressor did not produce consistent effects. The study contributes to IS research by highlighting the complexity of social meaning attribution in human–AI interactions and by informing the design and governance of responsible workplace technologies, including avatar-based training and bias-aware communication systems.

Share

COinS
 
Jun 14th, 12:00 AM

When Avatars Offend: Emotional and Behavioral Responses To Microaggressions In Human–AI Interaction At Work

Employee perceptions of microaggressions in AI-mediated workplace interactions, including those involving avatars, remain underexplored. This study examines how individuals perceive and respond to microaggressions perpetrated by humans versus avatars in hybrid work environments. We propose a controlled laboratory experiment in which participants interact with human and avatar aggressors of different genders, while collecting multimodal behavioral data (facial expressions, gaze, posture) and self-reported emotional and cognitive responses. Contrary to expectations, no significant differences emerged between human- and avatar-delivered microaggressions regarding emotional or physiological responses. Participants reacted similarly across conditions, and the gender of the aggressor did not produce consistent effects. The study contributes to IS research by highlighting the complexity of social meaning attribution in human–AI interactions and by informing the design and governance of responsible workplace technologies, including avatar-based training and bias-aware communication systems.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.