THE IMPACT OF ACADEMIC ARTIFICIAL INTELLIGENCE DEPENDENCY ON ACADEMIC INTEGRITY AND SELF - EFFICACY IN RESEARCH STUDENTS

Abstract

Background: The expanding integration of Artificial Intelligence (AI) in education has revolutionized academic practices, improving efficiency in research and writing but simultaneously raising concerns about academic integrity and reduced self-efficacy. In developing contexts such as South Asia, where digital literacy and ethical frameworks are still evolving, understanding these dynamics is crucial to ensure that AI complements rather than compromises educational integrity.

Objective: This study aimed to examine the relationship between AI dependency, academic integrity, and self-efficacy among research students, while identifying gender-based and demographic variations influencing these variables.

Methods: A quantitative cross-sectional study was conducted among 302 actively enrolled research students aged 18–50 years, selected through purposive sampling. Data were collected via three standardized instruments: the Students’ AI Dependency Questionnaire (α = .89–.91), McCabe/ICAI Academic Misconduct Inventory (α = .90), and the Academic Self-Efficacy Scale (α = .87). Descriptive statistics, Pearson correlation, linear regression, and independent t-tests were applied using SPSS v27 to assess associations and group differences at a significance level of p < 0.05.

Results: The mean AI dependency score was 91.24 (SD = 16.4), academic integrity 33.66 (SD = 13.82), and self-efficacy 40.67 (SD = 8.63). AI dependency positively predicted academic misconduct (β = .14, p = .019, R² = .42). Academic integrity was negatively correlated with self-efficacy (r = –.20, p < .01) but positively correlated with AI dependency (r = .14, p < .05). Male students showed higher AI dependency (M = 94.68 ± 19.22) than females (M = 89.70 ± 14.22, p = .032).

Conclusion: Findings indicate that increased reliance on AI tools may enhance confidence yet elevate risks of academic misconduct, highlighting a paradox between technological competence and ethical vulnerability. Strengthened AI literacy and institutional ethics policies are vital to promote balanced, responsible AI use and preserve academic integrity.