ARTIFICIAL INTELLIGENCE AND SOCIOLOGY: TOWARD A SOCIOLOGICAL APPROACH TO UNDERSTANDING INTELLIGENT TECHNOLOGIES, INTELLIGENCE ARTIFICIELLE ET SOCIOLOGIE : VERS UNE APPROCHE SOCIOLOGIQUE POUR COMPRENDRE LES TECHNOLOGIES INTELLIGENTES

Abstract

Artificial intelligence (AI) is often described as a technical capacity to automate prediction, classification, generation, and decision-making. A sociological perspective, however, asks how these capacities are embedded in institutions, social relations, economic interests, cultural meanings, and unequal distributions of power. This article develops a conceptual framework for studying intelligent technologies as socio-technical arrangements rather than autonomous tools. It combines insights from classical sociology, the sociology of technology, critical data studies, and AI ethics to examine five connected processes: datafication, algorithmic classification, institutional delegation, labour reorganisation, and the production of trust and legitimacy.

The article argues that AI does not simply enter a pre-existing society; its design and use reorganise social practices while being shaped by existing hierarchies, infrastructures, and norms. The consequences are therefore neither inherently emancipatory nor inevitably oppressive. They depend on data quality, institutional accountability, user participation, ownership structures, and the capacity of affected communities to contest automated outcomes. A comparative analytical table and a proposed research model identify variables suitable for empirical investigation. The discussion also highlights the relevance of a sociological approach in Arabic-speaking societies, where language representation, uneven digital access, public-sector transformation, and imported technological systems deserve particular attention. The article concludes with research and policy priorities for socially accountable AI, including participatory impact assessment, algorithmic transparency, independent evaluation, and sustained public digital literacy.