Artificial Narrow Intelligence and Organizational Effectiveness in Higher Education: The Parallel Mediating Roles of Technology Acceptance and Organizational Culture

Artificial narrow intelligence (ANI) is being embedded into teaching, research administration, and governance faster than management theory can evaluate it, yet empirical evidence on how ANI shapes organizational effectiveness in higher education remains fragmented. This study proposes and tests the ANI Dual-Pathway Framework, in which technology acceptance (TA) and organizational culture (OC) act as parallel mediators linking ANI to the organizational effectiveness (OE) of academic staff. A sequential quantitative-qualitative design combined a structured questionnaire (511 valid responses from 550 distributed; 97.5% effective recovery rate) with 20 semi-structured interviews and focus-group discussions across public, private, and vocational higher education institutions in Beijing. Confirmatory factor analysis and structural equation modeling, estimated with maximum likelihood and 5000-resample bias-corrected bootstrap for specific indirect effects, were applied. The measurement model was reliable and valid (Cronbach's alpha = 0.914-0.952; KMO = 0.931; CFI = 0.994; RMSEA = 0.015), and the respecified structural model fit well (χ²/df = 1.691; CFI = 0.991; RMSEA = 0.037). ANI had a significant direct effect on OE (β= 0.290). The bootstrap specific indirect effects showed significant indirect effects through TA (β= 0.061) and through OC (β= 0.097). The organizational-culture indirect pathway exhibited a larger effect magnitude relative to the technology-acceptance pathway. Two exploratory propositions regarding socio-technical strain and AI anxiety were examined exclusively using qualitative interview data and were not subjected to quantitative hypothesis testing. Findings indicate that organizational culture constitutes the stronger mediating channel and that investments in trust building and benevolent leadership should accompany technical deployment. Given the cross-sectional nature and geographically bounded sample, causal inferences cannot be definitively established. This work provides evidence-based guidance for higher education managers and policy makers in emerging-economy contexts.