Validating Institutional AI Tool Adoption: Real-World Barriers and Determinants

The present paper is the findings of Phase 3 of a mixed-method research on the adoption of AI tools by Internal Quality Assurance Cells (IQACs) in higher education institutions in India. Based on the outcomes of Phase 2 methodology, Phase 3 will focus on the actual data collection developments (50% done in 25 institutions) and confirmation of research hypotheses. Using the identical convergent parallel mixed-methods design of 50 IQAC coordinators in five regions of India, the study confirms four Phase 2 results: (1) Perceived Usefulness as the most important adoption determinant, (2) Digital Readiness as an essential enabling factor, (3) Institution Type Variation in adoption rates, and (4) Adoption as a progressive 18-month process. Three barriers emerge out of the real world data collection, which include inconsistency of data format between the regulatory bodies, low technical competency of personnel, and poor IT infrastructure. Regional analysis indicates that Gujarat (15 institutions) is the most progressive in the initiatives of adoption whereas Maharashtra (7 institutions) and others are lagging. Phase 2 quantitative predictions are confirmed by qualitative evidence, and systemic barriers that are beyond technological limitations are identified. Implications deal with IQAC leaders, policymakers and the higher education systems. This validation study offers empirical data that implementing AI tools effectively necessitates systemic change in data governance, investment in infrastructure, and capacity-building of staff, rather than technology itself.