Abstract:

Artificial intelligence (AI) is rapidly transforming higher education institutions, yet its integration into the external quality assurance (QA) processes that govern them remains limited and contested. While existing discussions largely focus on AI in teaching and learning, much less is known about how evaluation organisations themselves integrate AI into their practices. This paper addresses this gap by examining how AI is operationalised within evaluation processes and how agencies respond to its potential to reconfigure evaluative authority. The analysis develops a framework that conceptualises AI integration along three dimensions: as an instrumental tool supporting evaluation tasks, as a procedural object embedded in formal guidelines and protocols, and as a component that interacts with and potentially reconfigures evaluative authority. Empirically, the paper draws on policy and strategic documents from three QA agencies: the Norwegian Agency for Quality Assurance in Education (NOKUT), Quality and Qualifications Ireland (QQI), and the Australian Tertiary Education Quality and Standards Agency (TEQSA). The findings show that AI is not simply adopted as a neutral technological enhancement but is institutionally positioned in distinct ways — technological enclosure, branding defence, and regulatory expansion. Despite these differences in strategy, all three agencies converge on a common outcome: the preservation of human evaluative authority at the core of formal quality judgement. The paper argues that the significance of AI lies not in replacing human evaluators but in reconfiguring evaluation as a hybrid practice that combines algorithmic assistance with carefully maintained institutional control.

Keywords:

Artificial Intelligence, Machine Learning, Quality Assurance, Higher Education, Evaluation Practices.