Generative AI in Programming Education: Ethical Boundaries Across Five University Sites

Abstract

Aim/Purpose: This study examines how university students judge the ethicality and perceived institutional rule violation of generative artificial intelligence (GenAI) use in programming coursework and whether these judgments vary with the amount of AI-generated code, demonstrated understanding, stage of use, and university site.Background: GenAI can support explanation, debugging, and personalized programming learning, but it can also perform work that assignments are intended to use as evidence of student competence. IT educators therefore need evidence about where students locate the boundary between learning support and unacceptable substitution.Methodology: A scenario-based survey was completed by 195 students at five university sites in Botswana, Brazil, Canada, Indonesia, and South Korea. Participants evaluated eight programming scenarios using two five-point prompts addressing whether the conduct was unethical and whether it was against institutional rules. Data were analyzed using descriptive statistics, Kruskal-Wallis tests, one-vs-rest Mann-Whitney U contrasts, and site-specific ordinal logistic regression.Contribution: The study integrates Jones’s issue-contingent perspective with Treviño’s person-situation interactionist perspective to frame GenAI ethical judgment in programming education. It distinguishes students’ personal ethical judgments from their perceptions of institutional rules and treats differences among university sites as contextual variation rather than evidence of national culture.Findings: Responses differed significantly across the five sites for every scenario and both prompts. Across all sites, the amount of AI-generated code incorporated into the final submission was the only consistently positive and statistically significant scenario predictor. Demonstrated understanding and stage of use showed much smaller and generally non-significant associations. Ethical judgments and perceived institutional rule violation followed similar but non-identical patterns.Recommendations for Practitioners: Programming educators should state explicitly how GenAI may be used for explanation, debugging, code generation, revision, and submission; design assessments that make student reasoning visible; and use concrete scenarios to discuss acceptable GenAI use before graded work.Recommendations for Researchers: Future studies should employ factorial vignette designs, participant-level mixed-effects models, and direct measures of policy clarity, AI literacy, prior GenAI exposure, and assessment context. Comparisons between student and instructor judgments are also needed.Impact on Society: Clearer educational boundaries can help preserve authentic learning while enabling students to develop responsible GenAI practices that are increasingly relevant to professional computing work.Future Research: Further research should test the framework with larger and more diverse samples, examine actual GenAI behavior rather than hypothetical judgments, and extend the scenarios to other computing tasks and disciplines.