Artificial intelligence in service‐learning: A systematic mapping review and future research agenda for higher education

Abstract The rapid integration of artificial intelligence (AI) into higher education has begun to transform service‐learning (SL) by creating new opportunities for experiential, data‐informed and socially responsive pedagogy. However, the nature of AI integration in SL and its implications for learning and community engagement remain unclear. Guided by the PRISMA protocol, this systematic mapping review synthesised 26 peer‐reviewed studies screened and managed using EPPI Reviewer. The findings show that direct service and onsite delivery emerged as the dominant configurations, while AI was used most frequently for data analysis and interpretation, content generation and co‐creation and adaptive support and personalisation. Across SL phases, AI integration clustered around preparation and action, with considerably less attention to reflection. Reported outcomes centred primarily on knowledge application, transferable skills, civic orientation and personal development, whereas beneficiary outcomes appeared in only two studies. Alongside these outcomes, the studies identified concerns involving output unreliability, over‐reliance, AI readiness, ethical safeguards, contextual fit and access. Fewer than half used a recognisable theory or established framework, and none tested or refined theory. Overall, the review identifies an emerging field in which technological experimentation has advanced faster than evidence concerning AI‐specific contribution, community value, sustained pedagogical integration and theoretical explanation. Context and implications Rationale for this study: Artificial intelligence is increasingly being incorporated into service‐learning (SL) across higher education, but the emerging literature provides a fragmented account of how this integration is taking place. Existing studies span different service configurations, AI functions, phases of the SL process and disciplinary settings, making it difficult to understand the direction in which the field is developing. This study systematically maps this evidence to clarify how AI is being integrated into SL, what outcomes and concerns have been reported and how theory has informed this emerging area of research. Why the new findings matter: The findings matter because they challenge the assumption that greater use of AI necessarily transforms the fundamental character of SL. Despite growing technological integration, direct and onsite community engagement continues to anchor much of the experience, while AI is concentrated in particular functions and phases of the SL process. The findings also expose important weaknesses in the emerging evidence base, particularly the limited attention to reflection, community beneficiaries, AI‐specific contributions and theoretical explanation. By bringing these issues together, the review clarifies where AI is genuinely extending SL and where technological adoption is advancing without equivalent pedagogical or conceptual development. Implications for researchers and educational institutions: The findings point to a future research agenda centred on sustained AI integration across the SL process, clearer evidence of AI‐specific effects, stronger assessment of community outcomes, and deeper theoretical explanation. Researchers should examine how particular AI functions interact with different SL phases and whether their contribution extends beyond outcomes already associated with conventional SL. For educational institutions, the findings highlight the importance of integrating AI in ways that preserve human judgement, meaningful community engagement and reciprocal value while addressing readiness, ethical safeguards and access.