Purpose Artificial intelligence (AI) is fundamentally reconfiguring knowledge practices, yet its effective integration into existing knowledge management (KM) infrastructures remains complex. This paper aims to address: How do organisations meaningfully integrate AI into operational practices in light of existing KM? This study moves beyond descriptions of individual AI tools to develop a holistic, empirically grounded theory of AI-driven knowledge transformation. Design/methodology/approach Using a rigorous, qualitative, interpretive grounded theory approach, this study analysed 100 publicly available AI implementation case studies. Initially, open codes were created through line-by-line coding and then systematically abstracted into coherent axial codes, which in turn informed the final selective codes. Findings Our analysis reveals a novel theoretical framework that posits useful AI integration in four distinct archetypes, shaped by three primary organisational decisions: externalise selective expert knowledge (vertical vs horizontal), consolidate knowledge into stable artefacts (verified vs non-verified) and workflow configuration (automation vs augmentation). Research limitations/implications This study is limited by its reliance on publicly available success stories, which may bias findings towards positive outcomes. Future research should examine instances of AI implementation failure to provide a more balanced perspective. Furthermore, the constructs developed from this secondary data require empirical validation through primary data collection. Originality/value This paper’s primary contribution is its novel, empirically grounded typology of AI integration strategies. These frameworks offer insight into how KM–AI integration is implemented in practice. It moves beyond abstract principles to offer a structured, actionable model for meaningful organisational transformation in the age of AI.
Balage et al. (Fri,) studied this question.