GenAI and the Transformation of Cognitive Production: A Regime Perspective

Generative artificial intelligence (GenAI) is rapidly reshaping IT-intensive knowledge work and has become a central concern for Information Systems (IS) research. While existing studies document productivity effects, new forms of human–AI collaboration, and emerging governance frameworks, they primarily examine adoption dynamics and oversight mechanisms. Less attention has been devoted to clarifying how GenAI reorganizes cognitive production itself. This theory-development paper advances a regime-level perspective to address this gap. We argue that by embedding execution-level cognition within technical infrastructures, generative systems expand the delegability of routine knowledge work and thereby reorganize abstraction, authority, and accountability across organizational layers. We conceptualize this configuration as cognitive Taylorism, a mode of organizing work in which routine expertise is progressively capitalized through infrastructure, execution and integrative judgment are partially decoupled, and professional career pathways compress. Drawing on labor process theory as an analytical lens, we develop a structural account of how this regime reshapes organizational layering, redistributes expertise, alters authorship and identity, and restructures capability formation in IT-related occupations. In doing so, our paper reframes GenAI as a transformation in how cognition is organized and outlines a research agenda that positions workforce dynamics, professional development, and institutional reproduction as central domains for IS scholarship in the age of generative infrastructures.