Beyond access: AI literacy, instructional design and perceived autonomous learning among students in Indian higher education

Abstract

Universities are rapidly adopting artificial intelligence, yet access to AI tools does not guarantee that students experience greater ownership of their learning. This study examines how three learner side factors, AI literacy, AI readiness and perceived value of AI tools, and one pedagogy side factor, AI enhanced instructional design, relate to perceived autonomous learning outcomes. Drawing on the Technology Acceptance Model and Self Determination Theory, we conducted a cross sectional online survey of students in Indian higher education who had used AI supported tools for academic purposes in the previous six months. After data cleaning, 516 valid responses were analysed. The measurement model showed strong reliability and convergent validity, and discriminant validity was supported (all HTMT values below 0.85). A simultaneous multiple regression explained 46.7 percent of the variance in perceived autonomous learning outcomes (F(4, 511) = 111.887, p < .001). All four factors were positively associated with the outcome, with AI literacy the strongest (β = .336), followed by AI enhanced instructional design (β = .178), perceived value of AI tools (β = .147) and AI readiness (β = .120). Because the data are self reported, cross sectional and drawn from a nonprobability sample, the results indicate associations among perceptions, not causal effects or objective learning gains. The findings suggest that institutions should pair AI adoption with systematic AI literacy development and thoughtful instructional design, treating student agency rather than tool availability as the goal.