Abstract
As generative artificial intelligence (AI) becomes embedded in higher education, universities are redesigning assessment and policies to support responsible AI use while maintaining academic integrity (Nguyen et al., 2025; Weng et al., 2024; Xia et al., 2024; Zhao et al., 2024). Emerging scholarship has established key principles for responsible AI use (Nguyen et al., 2025), and translated these into institutional policies, assessment guidance, and disclosure requirements that shape expectations for student AI use (Furze et al., 2024; Humble, 2025; Ucan & Demirel Ucan, 2026). While this work has advanced understanding of responsible AI use in higher education, much less is known about how students understand these expectations once they are embedded in assessment practice.
One increasingly common mechanism for communicating these expectations is through compulsory AI declarations, which require students to disclose how they have used AI in completing assessment tasks. Research has begun examining AI declarations as mechanisms for promoting transparency, academic integrity, and student compliance (Gonsalves, 2025), while also proposing more structured declaration frameworks to support responsible AI use (Micallef & Petrovska, 2026). However, AI declarations have rarely been examined as student-authored texts through which students interpret and communicate expectations for responsible AI use.
This study examines compulsory AI declarations submitted as part of a first-year interdisciplinary business course assessment with approximately 500 students enrolled in each offering. Students are permitted to use AI only during the planning stage of a written consulting report and must submit an accompanying AI declaration. Rather than treating these declarations simply as records of AI use, this study views them as student-authored texts produced in response to an institutional requirement for responsible AI disclosure. Accordingly, it asks: How do students construct accounts of responsible AI use through compulsory AI declarations?
To address this question, the Academic Literacies perspective (Lea & Street, 2006; Lillis, 2003), which conceptualises student writing as a socially situated practice shaped by institutional expectations, provides the conceptual foundation for the research. Accordingly, AI declarations are examined not simply as records of AI use, but as texts that reflect how students make sense of responsible AI use within their assessment context. Consistent with this perspective, qualitative discourse analysis is adopted because it examines language-in-use and how meaning is constructed and communicated within specific educational contexts, rather than treating texts as neutral records of information (Imafuku et al., 2022). The analysis will be conducted in accordance with institutional human research ethics requirements, and all declarations included in the research will be analysed in de-identified form.
The study examines the proposition that AI declarations extend beyond administrative disclosure. Specifically, it investigates how students use AI declarations to interpret and communicate their understanding of responsible AI use within assessment. By shifting attention from the design of AI policies and disclosure frameworks to students’ interpretations of these expectations, the analysis contributes a new perspective to emerging discussions of AI-assisted assessment. Findings will inform the design of AI declarations, assessment guidance, and institutional AI policy by highlighting how students make sense of, rather than simply comply with, expectations for responsible AI use.
Keywords
Sense making, AI declaration, Academic literacies, Responsible AI use.
