Year: 2026
Type: article
Abstract: The rapid diffusion of generative artificial intelligence (AI) challenges the validity of written assessment in legal education by weakening the relationship between submitted coursework and students’ individual legal reasoning. This article argues that generative AI should be understood not primarily as a problem of academic integrity, but as a stress test for assessment validity. Drawing on a practitioner-led, design-based study of a first year module addressing business law, it examines an assessment redesign that made AI use mandatory and transparent. Students were required to evaluate, challenge and substantiate AI outputs using authoritative legal sources, with marks awarded solely for evaluative judgment, legal reasoning and source use. The article conceptualises generative AI as an epistemic artefact that can foreground engagement with legal authority, provides an empirical account of AI-inclusive formative and summative assessment, and proposes a transferable model informed by constructive alignment, transparency and ethical engagement. The findings suggest that regulated AI use can support the assessment of legal reasoning while reducing integrity concerns. The central challenge posed by generative AI, therefore, concerns what assessments can validly measure rather than misconduct alone.
Source: The Law Teacher
Author: Fernando Barrio
Institution: Queen Mary University of London
Language: English
Subfield: Law
Field: Social Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7225902324
