Abstract

The increasing integration of artificial intelligence (AI) in higher education highlights the growing importance of AI literacy among students and educators. AI literacy extends beyond technical proficiency and includes critical thinking, ethical awareness, and the ability to evaluate AI-generated content. Despite its significance, limited research investigates how students perceive AI tools, particularly in resource-constrained and policy-fragmented contexts like South Africa, and what institutional support they require for responsible AI use. This study employs a qualitative research design to analyze survey responses from seventy-eight students, using thematic analysis aligned with Zhou and Schofield’s AI literacy framework. Findings reveal that, while students recognize AI’s potential to enhance learning, they struggle with unclear institutional policies, unequal access to digital resources, and uncertainty about academic integrity when using AI tools. Many students expressed confusion over what constitutes acceptable AI use and called for structured AI literacy programs, clearer guidelines, and active faculty engagement in AI-related training. By applying a globally recognized framework to an underexplored setting, this study offers a localized interpretation of global AI literacy debates, demonstrating how digital inequality, multilingualism, and fragmented policy environments shape student experiences in ways not adequately captured by existing literature. The study recommends that universities embed AI literacy into curricula, establish clear institutional policies on AI-assisted learning, and enhance access to digital resources to foster equitable and responsible AI adoption. These findings contribute to global AI literacy discourse while advancing context-sensitive strategies that respond directly to the realities of South African higher education.