Generative AI in Social Enterprises: A Scoping Review of Research Trends and an Agenda for Korea

Generative artificial intelligence is entering social-enterprise and nonprofit work, yet evaluating its organizational implications requires evidence that distinguishes expectations from observed outcomes. This scoping review maps AI and digital-transformation literature published from 2015 through 11 November 2026 and identifies a 2022–2026 generative-AI subset. Guided by a local protocol informed by Arksey and O’Malley and PRISMA-ScR, the review conducted eight OpenAlex title-and-abstract searches and supplementary Korean KCI and RISS searches. Retrieval yielded 16,390 OpenAlex record-query associations and 55 Korean-portal rows; deduplication left 15,343 unique records. AI-assisted selection and source verification retained 1,299 publications, including 41 addressing generative AI. The publication-level map distinguishes organizational accounts, prospective student studies, communication experiments, research applications of machine learning, conceptual models, reviews, and teaching cases. Reported uses span communication, service delivery, fundraising, governance support, and platform-based business models. However, survey perceptions, intentions, scenario ratings, and theoretical benefits do not establish a common causal effect on organizational performance or social impact. Korean digital-transformation studies offer relevant capability and expectation constructs, though the national setting of each source must be verified independently. Seven proposed research questions address population definition, sustained capability, task-level net effects, beneficiary agency, decision authority, shared support, and donor trust. Accompanying search, extraction, and verification records support revision. The author’s manual audit of the inclusion list was not completed before release; this is an AI-assisted working paper, not an externally peer-reviewed article. Series: SEDI Working Paper No. 2026-01, Society for Social Enterprise and Digital Innovation (SEDI), Korea. Not externally peer reviewed. AI tools assisted with screening, drafting and copy-editing; see the Declarations section of the PDF. Files: manuscript (PDF), Korean abstract, extraction table, search log, screening audit, reference-verification registry, seeded audit sample and PRISMA flow diagram. Version 1.1 (11 November 2026): APA 7 formatting of version 1.0 (https://doi.org/10.5281/zenodo.24129144). Text, numbers, citations, references and companion files are unchanged; two section cross-references were reworded to match unnumbered APA headings.