The non-neutrality of algorithms and their use by digital platforms

Abstract

This article explains the algorithmic mediation of everyday activities in today's society by digital platforms. We analyzed the basic composition of an algorithmic code, which contains scripts that can reveal views shaped by the political and economic interests of its owner. To exemplify these scripts, we created two excerpts of similar deterministic algorithms, but which perform two opposing actions: one of exclusion and the other of inclusion of the low-income population in a hypothetical credit granting system. We found that this exclusionary/inclusive bias is part of the decision-making to be carried out by the code and eliminates any neutrality. In our investigation into the non-neutrality of algorithms, we conducted a literature review focused on the theory of platform capitalism, which provides us with insights into the mode of production based on the use of data and algorithms by digital platforms. Concluding that algorithms are not neutral, we propose, in a decolonial approach, a questioning of the biases of the algorithmic systems of the digital platforms we use, especially digital media platforms.