Navigating Ethical Boundaries: Educators’ Perspectives on Assessing Skills with Generative AI

Abstract

Paper presented at the Canadian Symposium on Academic Integrity 2025. The emergence of Generative Artificial Intelligence (GenAI), powered by deep learning models capable of producing human-like content (Lim et al., 2023), is transforming higher education by reshaping how knowledge is created and assessed in educational contexts. While scholars emphasize the need to explore GenAI’s ethical implications (Espartinez, 2024; Kooli, 2023), its integration into educational settings raises pressing concerns about academic integrity and assessment practices. Despite these concerns, there is a notable gap in empirical research on educators’ perspectives—critical voices for guiding GenAI’s responsible and effective integration into assessment (Bower et al., 2024; Nikolic et al., 2023). This study addresses this gap by examining how educators interpret the ethical boundaries of assessing student skills with GenAI, particularly distinguishing which skills can—and cannot—be evaluated using these tools. We present qualitative findings from interviews with twenty-eight post-secondary educators from Canada, England, and Australia. Using thematic narrative analysis (Merriam & Tisdell, 2016), guided by a teaching-and-learning lens to academic integrity, we analyzed how educators made meaning of GenAI’s integration into assessment practices. Findings reveal that participants viewed prompting and critical thinking as skills that could be meaningfully assessed in conjunction with GenAI, with prompting closely linked to communication and metacognitive abilities. Educators saw GenAI’s limitations—such as weak arguments and hallucinations—as opportunities to strengthen students’ evaluative judgment through assessments integrating GenAI. While some viewed GenAI as a co-pilot fostering creativity and expanding perspectives, others expressed concern that its use in assessing writing skills could encourage overreliance. These insights suggest that educators must recalibrate assessment designs to cultivate prompting and critical thinking skills while ensuring that foundational academic skills, such as writing and independent reasoning, continue to develop alongside GenAI’s integration into assessment. Recommended citation: Eaton, S. E., Moya Figueroa, B., Kumar, R., McDermott, B., Brennan, R., & Wiens, J. (2025, May 22–23). Navigating Ethical Boundaries: Educators’ Perspectives on Assessing Skills with Generative AI Canadian Symposium on Academic Integrity, University of Regina, Regina, SK. https://hdl.handle.net/1880/121796