Artificial Intelligence in Knowledge Management: Mapping a Decade of Research and Emerging Directions

This article is part of an ongoing series looking at AI in KM, and KM in AI.

• A science-mapping analysis was conducted on 209 English-language journal articles indexed in Scopus from 2015 to 2025. Consolidated bibliometric mapping combines performance indicators, collaboration patterns, and thematic clustering within a single analytical narrative.
• The findings reveal a sharp increase in KM–AI publications after 2020, indicating a transition from early exploratory studies to a rapidly expanding research domain.
• Social-structure analysis shows that knowledge production is globally distributed but concentrated among a core group of countries, institutions, and author networks, with collaboration influencing topic diffusion.
• Artificial intelligence (AI) increasingly functions as an enabler of KM by automating and augmenting knowledge discovery, interpretation, and dissemination.
• A relatively small set of recurring contributors appears prominently in the research literature, and publications are anchored in several national contexts, including the United Kingdom, Italy, India, and Romania.  That concentration is not neutral. It can shape which empirical settings are most frequently studied and which assumptions quietly become “standard” in the literature.
• The conceptual landscape can be summarised into three pillars:
  • A decision-and-analytics pillar positions AI as a mechanism for knowledge discovery and decision support.
  • A people-and-leadership pillar frames AI adoption as socio-technical change that shapes collaboration, knowledge-sharing routines, and innovative work behaviour.
  • Third, a governance-and-sustainability pillar emphasises that KM outcomes remain contingent on trust, risk management, and responsible knowledge processes, and that AI can either mitigate or intensify these issues depending on implementation choices.
• Taken together, the clustering supports the view that AI-enabled KM value creation depends on alignment between analytics capability, human readiness, and governance arrangements.
• From a managerial perspective, the evidence cautions against framing AI-enabled KM as a purely technical rollout. Leadership commitment matters because it affects prioritisation, resource allocation, and the removal of barriers that typically slow adoption.
• Training investments are equally consequential; without them, AI tools may be used superficially and fail to translate into sustained changes in knowledge-related behaviours.
• Trust remains central as well. When employees do not feel safe contributing knowledge – or do not trust what the system returns – platforms struggle to generate the contribution and reuse cycles that KM depends on.

Header image: Publication growth trend in KM–AI (2015–2025). Source: Gunawan et al., 2026).

References: