Generative AI in higher education: adoption, risks, and opportunities in the Indian subcontinent
Agostinho Sousa Pinto, António Abreu, Manuel Pérez Cota, Jerónimo Paiva, Md. Sakib Biswas, Anam Jamal Ansari, Niroj Dahal, Sindhu SinghPurpose
This study aims to investigate generative AI (GenAI) adoption in Indian Subcontinent higher education – a region marked by extreme socio-economic inequalities, infrastructural deficits and linguistic diversity – examining how perceived ease of use, usefulness, satisfaction, trust and security influence students’ continuance intention.
Design/methodology/approach
A sequential mixed-methods design was used. Quantitative data from an online survey of higher education students (n = 704) across seven South Asian countries were analyzed using partial least squares-structural equation modelling. Qualitative data were collected via semi-structured interviews with 14 professors from six countries within the Indian Subcontinent. The Technology Continuance Theory (TCT) served as the analytical framework.
Findings
Perceived ease of use, usefulness, satisfaction, trust and security significantly shape continuance intention. Satisfaction emerged as the strongest predictor of continuance intention (ß = 0.571, p < 0.001). Applications concentrate on content creation, administrative automation and performance enhancement. However, limited broadband access, chronic power instability, entrenched inequalities and underrepresentation of regional languages in large language models constrain equitable adoption.
Research limitations/implications
The cross-sectional design limits causal inferences. Self-reported data may introduce bias and the sample overrepresents urban, English-proficient populations. Future research should use longitudinal designs, include administrator perspectives, experimentally evaluate interventions for linguistic barriers and conduct comparative studies across Global South and North contexts.
Practical implications
South Asian higher education institutions must address infrastructure deficits (reliable electricity, affordable internet), integrate GenAI tools with clear ethical guidelines, advocate for regional-language large language models, negotiate subsidized licenses and provide faculty development programs to bridge the generational digital divide.
Social implications
Without inclusive digital policies and culturally attuned regulations, GenAI risks exacerbating divides between elite urban universities and underfunded rural colleges and between English speakers and regional language users. North–South partnerships are essential for equitable and sustainable higher education in the Global South.
Originality/value
To the best of the authors’ knowledge, this study is among the first to examine GenAI continuance intention across the entire Indian Subcontinent (seven countries). It extends TCT by integrating region-specific barriers – infrastructural instability, linguistic exclusion and socio-economic inequality – alongside AI-specific antecedents. The mixed-methods design provides robust, context-sensitive evidence from both students and professors.