Artificial Intelligence and Academic Integrity in UK Higher Education: A Rapid Literature Review of Emerging Challenges, Institutional Policy Responses, and Future Research Directions

Abstract

Background: Academic integrity has been challenged in a radically different way in the UK higher education sector by the rapid growth of generative artificial intelligence (GenAI) tools, underscored by the widespread use of large language models, including ChatGPT. Just three years after ChatGPT was released in the UK in the end of 2022, adoption by students has jumped from about 53% to 88%, presenting new systemic challenges, for which the current institutional frameworks were not planned and are not well-suited to deal with. The method used in carrying out this study is a rapid literature review (RLR) with a literature reporting approach based on PRISMA 2020. Four academic databases (Scopus, Web of Science, ERIC and the British Education Index) were searched as well as policy databases in the UK such as the British Education Index's (BEI) Higher Education publications archive, and Jisc's National Centre on AI (NCAI) database. The Quality Assurance Agency (QAA) was also used as a source for policy documents related to artificial intelligence for education. A total of 38 sources were identified for consideration after systematic screening and included two institutional policy analyses, six official sector documents, and 29 peer-reviewed journal articles. The synthesis highlighted four key thematic areas: (1) the unprecedented scale and definitional complexity of academic misconduct involving AI; (2) the technical and ethical limitations of academic integrity detection systems for AI; (3) the different nature and weak cohesion of policy responses to academic misconduct across the UK sector; and (4) the general consensus on redesigning and authentic assessment and enhancing AI literacy as the educationally soundest strategic directions. As reviewed above, the scholarly literature at hand questions the conceptual soundness of current academic integrity frameworks, which are developed within a pre-AI world that relies on individual authorship and empirical evidence assessable through the traceability of student work. Along with this, the paper reveals four areas of research that require immediate attention: Longitudinal effectiveness studies; UK specific empirical studies based on student and staff perspectives; Comparative institutional policy analysis; Theoretical research in the creation of new academic integrity frameworks that are sufficient to the AI era. The subject aims to embed a deeper understanding of generative AI within the framework of academic integrity and the concept of academic misconduct. The aims of the topic are to build a more robust understanding of generative AI in the context of academic integrity and academic misconduct.