Implicit Sexist Bias in Language and its Impact on Artificial Intelligence

Abstract

In this article we discuss the possible sexist biases that arise in the use of gendermarked and unmarked linguistic structures in Spanish and their impact on the treatment of language in artificial intelligence. The first part analyses sexist biases that arise in explicit gender-marked linguistic structures and those triggered by pragmatic inferencing in unmarked structures. The second part explores the consequences that these biases may bring into the world of AI, especially in relation to tasks such as programming virtual assistants, machine translation, and sentiment analysis.

Keywords

• sexist communication
• gender-unmarked structures
• intentionality
• artificial intelligence
• spanish