Purpose: Higher education institutions face increasing pressure to audit course designs for generative AI (GenAI) integration. This paper presents an end-to-end method for using large language models (LLMs) to scan course information sheets at scale, identify where assessments may be vulnerable to student use of GenAI tools, validate system performance through iterative refinement, and operationalise results through direct stakeholder communication and effort. Method: We developed a four-phase pipeline: (0) manual pilot sampling, (1) iterative prompt engineering with multi-model comparison, (2) full production scan of 4,684 Bachelor and Master course information sheets (Academic Year 2024-2025) from the Vrije Universiteit Brussel (VUB) with automated report generation and email distribution to teaching teams (91.4% address-matched) using a three-tier risk taxonomy (Clear risk, Potential risk, Low risk), and (3) longitudinal re-scan of 4,675 sheets after the next catalogue release. Results: Five iterations of prompt refinement achieved 87% agreement with expert labels. GPT-4o was selected for production based on superior handling of ambiguous cases involving internships and practical components. The Year 1 scan classified 60.3% of courses as Clear risk, 15.2% as Potential risk, and 24.5% as Low risk. Year 2 comparison revealed substantial shifts in risk distributions, with improvements most pronounced in practice-oriented programmes. Implications: The method enables institutions to rapidly transform heterogeneous catalogue data into structured and actionable intelligence. The approach is transferable to other audit domains (sustainability, accessibility, pedagogical alignment) and provides a template for responsible LLM deployment in higher education governance.
Figures & tables
Iteration
Agreement
Primary Error Pattern
Key Prompt Change
1
63% (19/30)
Over-classification of internships
Clarified supervision criteria
2
70% (21/30)
Misweighting practical components
Added grade contribution guidance
3
77% (23/30)
Language inconsistency
Strengthened language requirements
4
83% (25/30)
Excessive caution
Refined Low risk criteria
5
87% (26/30)
Ambiguous boundary cases
(Production stable)
Table 1: Pilot phase iteration outcomes
Category
Count
Percentage
Clear risk
2,824
60.3%
Potential risk
713
15.2%
Low risk
1,147
24.5%
Total
4,684
100%
Table 2: Year 1 risk classification distribution (AY 2024–2025, all BA/MA courses)
Category
Count
Percentage
Clear risk
2,253
48.2%
Potential risk
1,024
21.9%
Low risk
1,398
29.9%
Total
4,675
100%
Table 3: Year 2 risk classification distribution (AY 2025–2026, all BA/MA courses)
Year 2
Year 1
Clear
Potential
Low
Total
Clear risk
1,847
412
156
2,415
Potential risk
189
298
156
643
Low risk
98
156
580
834
Total
2,134
866
892
3,892
Table 4: Category transition matrix (Year 1 to Year 2)
Large language model auditing tests models and deployed systems for bias, safety, privacy, reliability, and provenance using behavioral and internal evidence.
Student use of Generative AI (GenAI) products in completing their classwork, with or without their professors' knowledge and/or approval, has resulted in substantial shifts in higher education. While GenAI use is widespread, its impact on student study methods, faculty course development, grade reporting, and overall learning is not well documented. This is a mixed-methods, multi-course study using retrospective quantitative analysis, instructor surveys, and anonymous student surveys at a university in the New England region of the United States. This research seeks to identify and document patterns in student and faculty perceptions of, and experiences in, the use of LLMs as a learning tool inside and outside of the university classroom. Alongside quantitative and thematic analysis of both faculty and student survey responses, historical grade data as reported to the university registrar is used to triangulate the phenomenon of learning achievement in pre- and post-LLM eras. It is hoped that this research can serve as a pilot study for a broader set of institutions. Results from this study can inform GenAI policy for professors, universities, and other educational institutions that are trying to maximize student learning in the age of AI.
The rapid proliferation of large language models (LLMs) in the context of education has introduced significant challenges in enforcement of academic integrity, especially in programming courses. We present Argus, an automated detection system for LLM-assisted student work in undergraduate C programming assignments. Argus integrates behavioral and stylistic indicators to create a holistic picture of the student's progress through an assignment and surfaces anomalies that point to potential misuse of LLM assistance. We quantify and analyze data over six years of Spring semester offerings in a large-enrollment CS2 course at Purdue University using Argus, finding that 45% of enrolled students exhibited patterns consistent with LLM-assisted code development in Spring 2026. To contextualize these results, we analyze the relationship between flagged LLM use and student performance on written, in-person proctored examinations, and find a significant negative correlation. We also explore the problem of mitigating false positives, recognizing that erroneous accusations of academic integrity carry significant consequences for students and instructors alike, particularly in the context of large enrollment courses. We argue that any automated detection system must be accompanied by a structured process for human review. We discuss the consequences for future course design, changing academic policy as these tools become more ubiquitous, and the pedagogical implications of LLM-based tools in computer science education.
Generative AI and large language models (LLMs) are increasingly applied to question generation and automated assessment. However, deploying LLMs in preparation for high-stakes exams requires more than prompt engineering; it demands software pipelines that systematically ground model outputs in authorised curriculum artefacts and marking guidelines issued by education authorities. This paper presents a curriculum-grounded, configurable LLM-as-Judge pipeline for question-level marking, co-developed with an industrial partner, to support exam preparation for university admission. The pipeline identifies the relevant topics, subtopics, and cognitive demand of a question, and assembles verifiable and authorised context to support LLM judgement. Curriculum intent is operationalised through concrete syllabus artefacts, including prescribed verbs and outcomes, performance band descriptors, glossary definitions, and marking-guideline principles. A staged LLM workflow is employed to first generate question-specific rubrics, capturing structured expectations of performance, and then derive and evaluate marking criteria used to allocate marks to student responses. This design improves consistency, transparency, and alignment with official marking practices. Preliminary evaluation shows that the proposed LLM-as-Judge pipeline delivers marking outcomes comparable to human tutors, while yielding justifications that are more traceable to authorised curriculum artefacts and marking standards. The pipeline has also been integrated into an online study platform, where early deployment data provide initial insights into operational usage and manual overrides.
Xiwei Xu, Chen Wang, Jacky Jiang +5
CSIRO, Australia · UNSW, Sydney, Australia · Studitory, Australia +1