Why Students Use Generative Artificial Intelligence for Essay Editing: A Configurational Method and Multidimensional Ethics Scale Perspective
Background: The rapid emergence of generative artificial intelligence (GAI) is having a substantial impact across numerous areas of higher education, including academic writing. Its use as a linguistic assistant occupies an ethically ambiguous position between legitimate academic support and practices that may foster technological dependence or compromise academic integrity. Objective: Examining the use of GAI among university students in a specific academic scenario: its use as a linguistic assistant to improve the writing quality and clarity of essays previously produced by the student, without substantially altering their content. The objective is to analyze how different ethical judgments shape students’ adoption of this practice, which occupies an intermediate position between legitimate academic support and potential risks of technological dependence or academic fraud. Methods: The study draws on the Multidimensional Ethics Scale, considering four moral dimensions: justice, relativism, consequentialism, and deontology. It also incorporates sociodemographic and academic variables, including gender, employment status, and perceived academic performance. Methodologically, fuzzy-set qualitative comparative analysis is applied to identify causal configurations associated with both the use and nonuse of GAI. Results: Acceptance does not depend on a single ethical dimension but on specific combinations of moral judgments. Consequentialism emerges as the most relevant condition in the pathways leading to use and is often combined with favorable perceptions of justice and relativism. In contrast, rejection shows a more fragmented structure and lower explanatory coverage, particularly when unfavorable ethical assessments, especially consequentialist assessments, are combined with sociodemographic factors. Conclusion: Students’ acceptance and rejection of GAI-supported essay editing follow asymmetric configurational logics, showing how different ethical judgments combine to explain academic technology use in a morally ambiguous context. The findings highlight the need for clear institutional rules, ethical training, and transparent criteria for academic GAI use. Universities should distinguish between acceptable linguistic support and practices that may compromise academic integrity, while helping students develop responsible and reflective uses of GAI.