Type: article
Abstract: Drawing on actor–network theory, we analyze the integration of an AI system designed to support a manual activity grounded in tacit knowledge: agricultural harvesting. Based on a qualitative case study combining six months of participant observation, interviews, and analysis of system use and adjustments, we trace the trials of translation across the moments of problematization, interessement, enrolment, and mobilization. The findings show that stabilization remains fragile: a deficit of situated explainability, the invisibilization of human mediations, and centralized algorithmic governance constrain the network's capacity to consolidate. This article contributes to the sociotechnical literature on AI in organizations by identifying the configuration of "provisional stabilization" as a durable yet precarious organizational form in which the sociotechnical network functions without consolidating, persisting as long as minimal conditions of maintenance are met, and by specifying three interrelated mechanisms: situated explainability, human mediations, and algorithmic governance that condition its production and fragilization
Source: AIS Electronic Library (AISeL)
Authors: Mawadia, Anass, Ruchaud, Sabrina
Institution: Université de Poitiers
Language: English
Subfield: Management Information Systems
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7222992602
