Generative artificial intelligence and human language: The sociology of algorithmic suspicion of tools of plagiarism between bias and delusion of accuracy, and penetration of human language.

Abstract

This research belongs to the sociology of algorithms. It is a qualitative study that investigates the impact of programmes designed to detect AI-generated texts. The study argues that these programmes do not truly aim to safeguard academic research or ensure fairness among scholars; rather, they reproduce a dominant language and culture. They are shaped by linguistic and racial biases that push aside many non-native English speakers, especially those of Black or Latin American heritage. Trust in the algorithms that claim to expose AI content is, the study maintains, an illusion that conceals economic and symbolic interests. At the same time, it affirms that human language itself is being reshaped by an algorithmic habitus—a disposition formed through the algorithmic upbringing of generative artificial intelligence. This habitus now influences the ways of writing, thinking, and expressing ideas within human discourse. All of this calls for critical approaches and for the development of far-sighted Arab educational policies