Responsible AI integration in higher education: drivers and effects in the context of Bangladesh

Purpose This study investigates the factors influencing and the impacts of responsible AI integration in Bangladesh's higher education (HE). It evaluates how ethical and social considerations (ESC), perceived benefits (PB) and perceived threats (PT) affect attitude (ATT), and how technological knowledge (TK), content knowledge (CK) and pedagogical knowledge (PK) influence responsible AI integration in higher education (RIAIHE), which subsequently drives the overall impact of responsible AI integration (IRIAIHE). The measured outcomes include academic performance, career guidance, motivation, self-reliance and social integration. Design/methodology/approach Data were collected from 550 students and faculty at public, private and national universities across Bangladesh using a mixed-mode questionnaire. Participants represented a diverse range of genders, religions, educational backgrounds and disciplines. Partial least squares structural equation modeling (PLS-SEM) with SmartPLS 4 was used for data analysis. Findings Path analysis showed significant direct effects of ATT (p 0.001), ESC (p 0.001), PB (p 0.001) and TK (p 0.001) on RIAIHE and of RIAIHE on IRIAIHE (p 0.001). However, CK, PK and PT did not have significant direct effects. Indirect effects revealed that ESC and PB significantly influenced RIAIHE and IRIAIHE through ATT, while TK impacted IRIAIHE via RIAIHE. The study also confirmed that ATT serves as a key mediator linking upstream variables to RIAIHE and IRIAIHE. Research limitations/implications These findings emphasize the importance of attitude, developing technological skills, ethical awareness and a positive outlook on AI for promoting its responsible and effective use in higher education. Originality/value The research offers practical insights for educators, academic leaders and policymakers aiming to harness AI for meaningful educational advancement.