Artificial Intelligence and Entrepreneurship Within Academic Ecosystems: A Bibliometric Analysis (2019–2025)

Abstract

This article addresses a specific gap in the bibliometric literature on artificial intelligence and higher education.While prior reviews by Kavitha et al. (2024), Flores-Velásquez et al. (2024), and Sahar and Munawaroh (2025)have broadly mapped AI adoption in higher education, none has jointly examined AI's role within universityentrepreneurial ecosystems, nor systematically tracked the generative AI inflection point that emerged after 2022in this sub-field. To address this gap, the study applies a bibliometric protocol grounded in the PRISMA frameworkto an initial pool of 222 Scopus-indexed documents published between 2019 and 2025, reduced to a final corpusof 165 articles after applying disciplinary, temporal, and document-type exclusion criteria. The corpus wasprocessed with VOSviewer to map publication trends, leading journals, country contributions, and keyword cooccurrence networks. Results reveal a pronounced geographic concentration of research output, with China aloneaccounting for 37 percent of the corpus and the nine most productive countries jointly contributing nearly fourfifths of all publications. Keyword co-occurrence analysis identifies artificial intelligence, students, andentrepreneurship education as the field's dominant thematic anchors, with artificial intelligence acting as theprincipal structural connector bridging pedagogical and entrepreneurial sub-themes. These patterns point topedagogical innovation, entrepreneurial competency-building, and generative AI integration as the field's coreresearch clusters. From a managerial perspective, the findings call on university incubators and technology transferoffices to embed AI literacy and generative AI tools into entrepreneurship curricula and venture-support processes,consistent with the conceptualization of AI as a general-purpose technology. The study's main limitations stemfrom its exclusive reliance on the Scopus database and its purely quantitative bibliometric design; future researchshould combine multiple databases and incorporate qualitative approaches to further explore innovation dynamicswithin university entrepreneurial ecosystems.