ARTIFICIAL INTELLIGENCE AND ACADEMIC INTEGRITY: RETHINKING ASSESSMENT MODELS IN THE AGE OF GENERATIVE AI

Abstract

The research paper focuses on how generative artificial intelligence (GenAI) is disruptive to academic integrity and academic assessment practices in higher education. To deal with the critical lag between technological change and institutional control, the study uses a pragmatic mixed-methodology framework that is organized into three phases. Phase 1 will conduct a thematic document analysis of the academic integrity policies of major international universities (Harvard, Oxford, Melbourne, Delhi) in order to build a comparative legal and policy framework on AI-assisted authorship and disclosure. Phase 2 collects primary empirical evidence based on a structured survey (N = 500) of students and teachers in institutions with Management, Law, and Arts programs in Madhya Pradesh, India (Indore, Ujjain, Bhopal). The data is then thoroughly correlated and descriptively analyzed using IBM SPSS through modified versions of the Technology Acceptance Model (TAM) and Academic Dishonesty scales to visualize perceptions of behavioural intentions towards GenAI and perceived efficacy of plagiarism detectors. Phase 3 compares alternative assessment strategies (viva-based, project-based, and timed case analyses) with authenticity and AI misuse criteria. Summing up these stages, the work proposes the new model of academic integrity governance, stating that in order to create technological disruption-resilient assessment, the tripartite alignment of institutional regulation responses and the basic pedagogic reconstruction are needed. The results can allow the university administrators and curriculum developers to obtain empirical evidence and practical models that would assure pedagogical credibility in the AI-ubiquitous age.