Gender bias (in AI)

Abstract

In the face of the emergence of artificial intelligence systems in multiple spheres, the incorporation of a gender perspective is essential to prevent the perpetuation and strengthening of stereotypes and discriminatory behaviors through biases in algorithms. AI demands a conscious intervention in its development and algorithm formulation, along with adherence to key principles such as transparency and responsibility, vital to avoiding gender bias. Bias is demonstrated through various studies and entails discrimination. It generally originates from the underrepresentation of groups and tends to manifest indirectly. An analysis of the regulatory framework applicable to artificial intelligence systems is included, as far as it is relevant to the gender issue.