Generative AI in Higher Education: Academic Authorship, Validity, and the Transformation of Assessment

This paper analyses how generative artificial intelligence is reshaping the requirements of academic authorship in higher education and explores the implications this development holds for a contemporary culture of assessment. It begins from the observation that generative systems can substantially accelerate scholarly work in areas such as structuring, drafting, and tutorial support, while simultaneously generating new risks for academic integrity, validity, and reliability. To address this development, the paper develops the theoretical framework of functional tool indifference. What matters, on this view, is not the tool as such, but whether validity, traceability, and human responsibility are preserved throughout the scholarly process. On this basis, the paper proposes a distinction between a merely technical product and a human artefact. An AI-generated output becomes a human artefact attributable in scholarly terms only through the human act of validation – that is, through review, selection, revision, source checking, and responsible contextualization. From this perspective, academic authorship shifts away from mere text production towards capacities of validation, verification, and judgement. For higher education, this entails the need for a transformed assessment culture that evaluates not only end products, but also processes of production, reflective performance, and the logic of evidentiary support. The paper therefore argues for a dual model of assessment based on process integrity and artefact validity, and for sustained AI governance that strengthens higher education institutions as sites of methodological, reflective, and validation competence.