A Discourse Analysis of Interactions with Alexa Virtual Assistant Showing Reproductions of Gender Bias

Abstract

Virtual assistants are everywhere, and most of these artifacts come by default with a female name and voice. This article refers to Science, Technology and Society (STS) studies to analyze the reproduction of gender bias in the discourse of Alexa virtual assistant, making use of the French Discourse Analysis method. From the interaction with Alexa and comparison of its responses with those found in previous surveys, it is identified the discursive and ideological formations in such speech, aiming to find stereotypic patterns. By working on the identification of unconscious gender biases in the discourse, approaching HumanComputer Interaction, this study presents how the prejudice-laden patriarchal culture presents itself in mediation with technology. By analyzing the assistant’s answers to genderrelated interactions, an awareness can be developed to help with a strategy of technological solutions that adjust this experience and create a less harmful and more inclusive design