Abstract

Generative artificial intelligence (GenAI) has moved from a research curiosity to an everyday workplace tool in a remarkably short window. This paper reports an empirical study of how these technologies are reshaping knowledge management (KM) practices and, in turn, employee productivity. Drawing on survey data from 412 knowledge workers across technology, financial services, consulting and public-sector organisations, the study tests a mediated model linking GenAI adoption to productivity through knowledge-management effectiveness and perceived usefulness. Results from ordinary least squares regression and mediation analysis show a strong positive association between the intensity of GenAI use and self-reported productivity (β = 0.72, p < 0.001), with knowledge-management effectiveness partially mediating the relationship. Respondents reported the largest gains in drafting, information retrieval and document summarisation, while flagging persistent concerns around output accuracy, data privacy and the erosion of tacit expertise. The paper contributes fresh, current evidence to a fast-moving literature and offers practical guidance for organisations trying to capture the upside of GenAI without hollowing out their knowledge base.