Integrity Debt Audit
Brenden Taylor | MSCC | Summer 2026
Overview
This document features an example of artificial intelligence (AI) use in a classroom setting by Brenden Taylor, an Assistant Professor of Theatre at Motlow State Community College.
As deliberated in the subsequent pages of this document, the instructor used AI in the following capacities:
- To diagnose the likelihood of automation by AI in his course’s major assessments and then re-engineer the assignments to be more resilient to AI automation.
- To integrate ways of both featuring the strengths of AI and exposing its weaknesses, biases, and limitations into said assignments for the students’ own edification.
In their cumulative project reflection essays, many students admitted to being “intimidated yet impressed” by the project and its requirements on the front end and then reported a growth in appreciation of the content experienced therein.
Context
Motlow College’s Intro to Theatre course currently features an assignment in which students select a main character from a Shakespearean text and analyze that character’s presence in the script as if they were an actor preparing to portray them in a hypothetical live production.
The emerging problem with this assignment is AI use: many students simply paste or upload the assignment instructions into an AI tool, sometimes a dedicated essay generator, other times a general model like ChatGPT, Gemini, Grok, or Claude, and have it write the essay for them.
Textual analysis, interpersonal communication, and project management are core theatre industry skills, and the vulnerability of this project to AI automation makes authentic student efforts at growing those skills hard to verify. More than that, such vulnerability invites academic dishonesty and complicates fair grading, which often leads to faculty burnout and disenfranchisement.
It is with these problems in mind that the instructor set out to reconstruct the assignment to become more resilient to AI automation.
Application of AI
The instructor used AI as a tool in both instructional design and required project activity with in-class demonstrations of proper and improper AI use.
Instructional Design via the Integrity Debt Audit
At the onset of redesigning this assessment, the instructor sought the assistance of the Integrity Debt Audit, an application that pairs Google Gemini with a research-based framework to identify deficiencies in classroom assignments and offer suggestions for fortification against improper AI use.
This free application was developed by Dr. Sam Illingworth, Full Professor of Critical AI Literacy at Edinburgh Napier University. In his research, Dr. Illingworth has identified ten specific areas by which classroom assignments can be scored to assess their strength or vulnerability to AI automation.
The following aspects comprise the framework for the Integrity Debt Audit:
- Final product weighting
- Iterative documentation
- Contextual specificity
- Reflective criticality
- Temporal friction
- Multimodal evidence
- Explicit AI interrogation
- Real-time defense
- Social/collaborative labor
- Data recency
Other instructors seeking to test their assignments may simply upload their instructional briefs to the prompt box, and the application utilizes Google Gemini to apply Dr. Illingworth’s framework to the assignment in question.
The application then produces a PDF report of its findings on a scale of 0–50 points, with every aspect scaled up to 5 points each. Scores of 0–20 are deemed a “High” susceptibility rating, scores of 21–39 are deemed a “Medium” susceptibility rating, and scores of 40–50 are deemed a “Low” susceptibility rating.
To date, the original assignment consistently scores within the 10–20 range when assessed, which highlights the problems of this assignment as mentioned earlier.
The instructor used these findings to engineer a new way of assessing the students. Whereas the original assignment is a single essay submission, the new assignment is a portfolio that retains a simplified, more reflective version of the original essay while introducing small-group peer review and a scaffolded timeline of low-stakes deliverables which culminate into a live presentation of group findings with a whole-class, instructor-led Q&A defense.
With these new aspects now integrated into the assignment, the revision repeatedly scores within the 45–50 range.
An appendix featuring the original assignment instructions, the revision, and both vulnerability reports are readily available upon request.
New Project Requirement: Explicit AI Interrogation
Following Dr. Illingworth’s research and suggestions, the instructor incorporated a task in the revised project that shows students how limited AI is in interpreting live events, another core theatre skill.
In the new project, students must prompt a generative AI/LLM with a description of two favorite moments from their assigned production recording and specifically ask it to defend something about the performance that is not true. They must then compare the AI’s defense with their own opinions, discuss consistencies and inconsistencies, highlight the forced hallucination, and include screenshots of the conversation.
By highlighting the forced hallucination, students can see how easily AI can be manipulated and misused. Although small, simple, and almost arbitrary, it is the instructor’s hope that this task encourages students to recognize AI as a processing tool and not as a replacement for genuine human experience.
Outcomes
From the instructor’s perspective, the most appreciated outcome was students reporting positive impressions and engagement with both the content and the project itself. Many students reported enjoying reading the text in question and viewing their assigned recorded performance of it with their groups.
While some students still found ways to misuse AI for this project, their misuse was all the more evident to both the instructor and the rest of the class.
With minor tweaking, this assignment will be ready to deploy once again for the Fall 2026 semester.