Hallucinated Citations and Algorithmic Bias: Evaluating the Impact of Student AI Literacy on Epistemic Trust and Scientific Misinformation in Among Undergraduate Students in the University of Calabar, Nigeria

The increasing integration of generative artificial intelligence (AI) into higher education has raised concerns about hallucinated citations, algorithmic bias, epistemic trust, and scientific misinformation. This study investigated student AI literacy, epistemic trust, and susceptibility to scientific misinformation among undergraduate students at the University of Calabar, Nigeria. Specifically, it examined the relationship between AI literacy and epistemic trust; the relationship between AI literacy and susceptibility to misinformation arising from hallucinated citations and algorithmic bias; the predictive influence of AI literacy, hallucinated-citation awareness, and algorithmic-bias awareness on epistemic trust and misinformation susceptibility; and differences in these variables by gender, AI usage, academic level, and faculty. A correlational survey design was adopted, involving 400 undergraduate students from different faculties and academic levels. Data were collected using an expert-validated structured questionnaire. A pilot test of 40 students produced Cronbach’s alpha coefficients ranging from 0.78 to 0.91. Data were analysed using Pearson Product-Moment Correlation, multiple linear regression, independent-samples t-test, and one-way ANOVA at the 0.05 significance level. Findings revealed significant negative relationships between AI literacy and epistemic trust (r = −0.46, p < .001) and between AI literacy and misinformation susceptibility (r = −0.52, p < .001). Regression analyses showed significant predictive effects, with AI literacy the strongest predictor. No significant gender differences were found, whereas significant differences occurred by AI usage, academic level, and faculty. The study recommends four measures: institutional AI-literacy programmes; curriculum-integrated verification and citation training; faculty-specific AI-literacy interventions; and university guidelines for responsible AI use, source verification, and academic citation.