Answers are grounded in this paper and cite their sources.
Concern has grown that students’ reliance on generative artificial intelligence (AI), and its use for academic dishonesty, may weaken self-efficacy, motivation, and academic achievement, yet evidence on these relationships is scattered across inconsistent labels. Following the Arksey and O’Malley framework and the PRISMA extension for scoping reviews (PRISMA-ScR), this review mapped studies published between 2015 and 2025 that examined AI dependence or AI-related academic integrity alongside motivational constructs or academic achievement among higher education students. A Scopus search returned 172 records, of which 21 met the eligibility criteria and were charted. No eligible study predated 2024, eight were set in East Asia and none in Africa, and sixteen were cross-sectional surveys. The apparent inconsistency in the direction of the self-efficacy effect largely disappears once academic self-efficacy, which tended to protect against dependence, is distinguished from AI-specific self-efficacy, which tended to promote it. Dependence was associated with inflated self-efficacy, procrastination, and weaker learning, while associations with achievement were mixed. Amotivation and degree apathy predicted AI-assisted misconduct more strongly than AI use itself. The evidence base remains recent, geographically narrow, reliant on self-report, and almost without longitudinal or intervention research.
No takes yet. Share an insight, caveat, or question.
Ahmed et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: