AI anxiety among university librarians in Bangladesh: A multidimensional assessment using the Artificial Intelligence Anxiety Scale

Abstract

As artificial intelligence (AI) becomes increasingly integrated into library services, understanding librarians’ psychological responses to this technological transformation has become an important research priority. This study investigates AI anxiety among university librarians in Bangladesh and examines whether anxiety levels vary across demographic and professional characteristics. A cross-sectional survey was conducted using the Artificial Intelligence Anxiety Scale (AIAS). Data were collected from librarians employed in public and private universities in Bangladesh through convenience sampling. Following data-quality screening, 138 valid responses were retained for analysis. Descriptive statistics, reliability analysis, correlation analysis, and one-way analysis of variance (ANOVA) with Bonferroni post hoc tests were employed to examine patterns of AI anxiety and group differences. Overall, university librarians reported a moderate level of AI anxiety (M = 3.00). Anxiety levels were broadly comparable across dimensions, although Sociotechnical Blindness recorded a slightly higher mean score than the remaining dimensions. Educational qualification and professional experience showed limited overall associations with AI anxiety, while gender, age, and professional designation demonstrated little variation. The findings suggest that concerns regarding governance, transparency, professional identity, and organizational readiness may be more salient than technological learning challenges. The study extends research on AI anxiety to a Global South library context and contributes evidence that AI anxiety among university librarians is well understood as an interrelated socio-professional and organizational phenomenon. The findings highlight the importance of professional development, institutional support, and responsible AI governance in academic libraries.