.png){kind=icon}
.png){kind=icon}
Return
Assessment validity in the age of generative artificial intelligence: a critical review
.png){kind=icon}
PRE
AI
R
J
DOI：10.1080/02602938.2026.2683046.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Generative artificial intelligence (GenAI) presents significant challenges to assessment validity in higher education. In response, universities worldwide have invested heavily in risk-mitigation strategies. However, such efforts often prove insufficient given the rapid and on-going development of GenAI capabilities, prompting sustained debate about how assessment in higher education can remain valid in this evolving landscape. At the same time, inconsistencies in assessment design and implementation have raised concerns regarding the assurance of learning. This study presents a critical review of the literature on assessment validity in the context of GenAI through a discourse-analytic examination of articles published in 10 leading higher education journals between 2023 and June 2025. Articles were selected using predefined inclusion criteria, and discourse analysis was employed to synthesise dominant narratives within the literature. Three main discourses were identified: (1) risk-oriented framings that emphasise academic integrity breaches, (2) rule-based approaches centred on detection and policy enforcement, and (3) design-focused approaches that advocate resilient assessment redesign. The findings offer conceptual and practical implications for assessment practice and institutional policy in higher education in the age of GenAI.
Keywords:
Generative artificial intelligence
GenAI
assessment validity
higher education
Journal
A
IF:
0
Papers:
72
Citations:
0
Organization
Cited Papers