Democratization or Digital Enclosure? Socio-Economic Disparities in Access to AI and Digital Learning Tools in Higher Education

Abstract

Generative Artificial Intelligence (GenAI) is widely heralded as an equalizer in higher education, yet emerging evidence suggests that tiered monetization and computational requirements may reproduce existing inequalities. Anchored in Pierre Bourdieu’s theory of cultural capital and Jan van Dijk’s digital divide framework, this study investigates socio-economic disparities in student adoption of GenAI tools. Utilizing a cross-sectional survey of university students (N=450) across public and private institutions, bivariate and hierarchical regression analyses were conducted. Findings indicate significant institutional (χ² = 44.18, p<.001) and income-based (χ² = 68.92, p<.001) divides in access to premium AI tiers. Furthermore, "Digital Cultural Capital" (prompt literacy and critical auditing) and paid subscription access independently predicted significant variance in academic self-efficacy (ΔR² = .234, p<.001). The study concludes that paywalled algorithmic tools risk instituting a modern digital enclosure, necessitating policy interventions such as institutional licensing consortia and standardized prompt literacy curricula.