Higher education has two layers of AI policy. Institutional policy lives in the academic integrity code, set by provosts and academic senates. Course-level policy lives in the syllabus, set by individual instructors. The institutional layer is being slowly updated for AI. The course layer is mostly left to faculty to figure out, often with little support. The higher ed AI Use Policy template I built is for that course layer.
This is the third of three companion posts on AI Use Agreements for different educational contexts. The elementary post covers K-5. The middle and high school post covers Grades 6-12. This one covers undergraduate and graduate courses.
What’s Inside the AI Use Policy
The document is eight pages, fully editable in Word, and designed to be modular. Each page serves a different copy-paste purpose. Faculty can drop pages into their syllabus, LMS, assignment instructions, or rubric without having to reformat the whole thing.
The first two pages are instructor guidance. They include an editable field for your institution’s academic integrity policy reference, a full reference table for the five AI use categories (Category 1: No AI Permitted, through Category 5: AI-Integrated Work), AI citation guidance with brief example formats for APA, MLA, and Chicago, and an editable discipline-specific examples table covering humanities, social sciences, STEM, business, health professions, education, law, computer science, arts, and graduate research.
Page three is a syllabus-ready AI Use Policy. This is the one faculty paste into their syllabus or course handbook. It covers course position on AI, relationship to institutional policy, the five categories, acceptable and unacceptable use, disclosure requirements, accuracy and verification, privacy and intellectual property (including a specific note on research and confidential data), accommodations, academic integrity, equity, discipline-specific expectations, and an instructor modeling statement.
The remaining pages serve as practical artifacts: an Assignment-Level AI Use Table that maps each course assignment to a category, an AI Use Disclosure Form for individual assignments, a Reflective AI Use Statement for major projects (theses, capstones, dissertations), a Privacy and Responsible Use Checklist, and a Student Acknowledgement Form with full signature fields.
Why the Choices Matter
Three design decisions distinguish this edition.
First, the five-category scale. The middle and high school edition uses four levels (0 through 3). The full Perkins, Roe, and Furze AI Assessment Scale has five categories, and Category 5 (AI-Integrated Work) genuinely fits higher ed contexts. Digital humanities projects, generative design courses, prompt engineering assignments, AI literacy seminars: these are real uses where AI is part of the design itself, not just a tool used to produce traditional work.
Second, the modular structure. Higher ed faculty don’t need a single all-in-one document. They need pages they can drop into different contexts: a syllabus paragraph, an assignment instruction, a rubric, an LMS announcement. The eight pages are built to be used independently or together. The Assignment-Level AI Use Table, for example, can be pasted into a syllabus or kept as a separate appendix.
The Reflective AI Use Statement is the third distinctive piece. It’s designed for thesis advisors, capstone supervisors, and faculty teaching writing-intensive courses. Students complete it once at the end of a major project, describing where AI shaped their methodology, what they accepted or rejected, and how they verified outputs. It’s a metacognitive document, not a checklist.
How to Use It
Use what fits. Faculty in writing-intensive courses tend to lean on the syllabus policy, the disclosure form, and the reflective statement. STEM departments lean more on the syllabus policy, the assignment table, and the acknowledgement form. Discipline norms shape which pages get the most traffic.
The document is built for course-level adoption, but a department or program can also adopt sections collectively. The five-category scale is particularly useful when department faculty want a shared vocabulary for talking about AI across courses.
Companion Editions
This is one of three editions. The Elementary edition (K-5) uses a simplified three-level system tailored to younger learners. The Middle and High School edition (Grades 6-12) uses a four-level scale with a focus on disclosure. There’s also a Master Document that combines all three editions in one file.
Download the Higher Education Edition
The full editable Word document is here: https://medkharbach.com/wp-content/uploads/2026/05/AI-Agreement-Higher-Ed-Version.docx
The work is shared under a Creative Commons license. Use it, adapt it, share it with colleagues. If you adapt it in ways that work better for your discipline or institution, I’d love to hear about it.
