Abstract:

The increased use of Artificial Intelligence (AI) in Higher Education is fundamentally reshaping the foundations of teaching, learning, and the assessment of knowledge acquisition across the global higher education landscape. In response to these developments, numerous institutions in South Africa have introduced AI governance frameworks intended to regulate and guide the ethical and responsible use of AI in academic environments. While these frameworks seek to provide institutional guidance, their practical application remains contested, particularly as AI systems are increasingly capable of generating, structuring, and refining written content in ways that closely resemble human authorship. This evolving capability raises important questions regarding the boundaries between human intellectual contribution and algorithmically generated output. As such, this study examines how institutional AI governance frameworks conceptualise authorship, responsibility, and acceptable AI use within higher education policy discourse.

Drawing on a qualitative research design informed by critical discourse analysis, the study analyses openly accessible institutional documents, including AI governance frameworks, academic integrity policies, and assessment guidelines from South African higher education institutions. The analysis seeks to identify how institutional policies frame the role of AI in academic work and the assumptions about knowledge production, authorship, and academic labour that underpin these frameworks. The findings indicate that institutional AI governance frameworks predominantly position AI as a supportive or assistive tool within the academic writing process, while simultaneously maintaining traditional understandings of individual authorship as the primary basis of academic responsibility. Policy discourses frequently emphasise the need for transparency and disclosure of AI use; however, they offer limited conceptual clarity on how AI-assisted contributions should be interpreted within existing notions of authorship and intellectual accountability.

As a result, a tension emerges between the increasingly collaborative realities of human–AI text production and institutional policy frameworks that remain grounded in models of academic work centred on individual intellectual labour. The study contributes to emerging debates on AI governance in higher education by highlighting a conceptual gap within current institutional responses to AI, particularly in relation to how authorship and intellectual contribution are defined and regulated in digitally mediated academic practices.

These findings have important implications for the development of more coherent and conceptually robust policy frameworks capable of addressing the complexities introduced by AI-assisted knowledge production in higher education.

Keywords:

Artificial Intelligence, higher education, academic authorship, AI governance, academic integrity, policy discourse.