Enterprise architecture as a knowledge management discipline: evolution, challenges and AI-enabled future

Purpose This article addresses the transition of enterprise architecture (EA) from static documentation to a dynamic domain of knowledge management (KM). In this context, we examine how EA tools, governance procedures and artifacts influence organizational learning and strategic alignment, drawing on existing literature. Design/methodology/approach To conduct this study, we adopted a hybrid methodology that combines integrated synthesis and a semi-systematic review, based on the typology proposed by Snyder (2019). Findings This study demonstrates that EA continues to face several challenges, such as knowledge fragmentation into silos, loss of tacit knowledge, semantic underspecification, tool-related constraints and governance misalignment. However, in the coming years, EA could evolve into an intelligent and adaptive information system thanks to emerging technologies, particularly artificial intelligence (AI). The article concludes with a research agenda aimed at enhancing EA using the capabilities offered by AI. This paradigm shift would enable EA to support ongoing business transformation and act as a strategic accelerator in complex environments. Originality/value This article advances EA research by proposing a three-dimensional framework (epistemological, technological and organizational) that positions EA as a dynamic knowledge management discipline. Key contributions include (1) mapping EA artifacts and governance to knowledge management processes (capture, formalization, sharing and reuse); (2) systematically linking five EA-KM barriers to specific AI capabilities and (3) proposing a testable research agenda for AI-augmented EA systems. This synthesis uniquely bridges the EA, KM and AI literature, transforming EA from static documentation to intelligent knowledge infrastructure.