Governance of Hybrid Academic Production: Authorship, Agency, and Algorithmic Capital in the AI Era

Abstract

The integration of generative artificial intelligence into academic research has largely been debated through a normative lens that opposes responsible use to misuse. This article argues that AI does more than raise ethical or operational dilemmas: it exposes a fundamental misalignment between how scientific knowledge is actually produced and the anthropocentric assumptions on which its governance mechanisms historically rest. Drawing on the Extended Mind and Extended Self perspectives, sociomateriality, Actor-Network Theory, and Bourdieu's theory of distinction, we reconceptualize academic production as a hybrid sociomaterial process in which humans and algorithmic systems operate interdependently in the generation, organization, and validation of knowledge. Within this process, we introduce the concept of algorithmic capital: an emerging form of symbolic distinction grounded not in access to AI but in the capacity to orchestrate it while rendering its mediation invisible, thereby preserving the appearance of human authorship. We then propose a multilevel, recursive governance framework that recognizes the distributed nature of agency, incorporates algorithmic systems as actors, and shifts the focus of governance from final products to the practices and relations that constitute knowledge production. Finally, we discuss implications for authorship, scientific evaluation, and researcher formation, arguing that AI reopens fundamental questions for organizational theory about how work is defined, how value is attributed, and how distinction is constructed in knowledge-intensive contexts.