Why university guidelines do not give you the answer for your assignment
The "Seoul National University AI Guidelines", established by Seoul National University (SNU) on January 1, 2026, describe themselves as voluntary principles for members of the university community. Each instructor is to determine what is permitted or prohibited and how use should be reported, according to the nature of the course[2]. Korea University's guidelines of March 15, 2023, already stated that the instructor of each course makes the final decision on whether AI is permitted[3]. The March 30, 2026, guidelines from the Korea University Undergraduate College Remote Education Center also require instructors to state their policies in the syllabus[4]. Yonsei University (May 2024) and the University of Seoul (May 2026) likewise tell students to check the policies in the syllabus or course plan[5][6].
The table below summarizes public documents reviewed on November 2, 2026. It does not include separate rules established by departments or colleges, and it does not cover other universities.
| University | Document | Date issued | Where to find the rules | What students must disclose |
|---|---|---|---|---|
| Seoul National University | AI Guidelines | January 2026 | Course policies in the syllabus and other course documents | As specified by the course |
| Seoul National University's Department of Computer Science and Engineering | AI Use Guidelines | October 2026 | The syllabus and instructions for each assignment | Tools, where they were used, and verification methods |
| Korea University | Guidelines for Using AI | March 2026 | AI use guidelines for each course | Tool names, when they were used, and scope of use |
| Yonsei University | Guidelines for Using Generative AI | May 2024 | The syllabus and course policies | Models, prompts, and platforms |
| University of Seoul | Guidelines for Using AI | May 2026 | The course plan | Source disclosure in the specified format |
| Myongji University | Guidelines for Using Generative AI | August 2026 | The level specified for each assignment in the course plan | Scope of use, key prompts, reference materials, and details of review and revision |
AI use levels differ by assignment, even within the same course
The guidelines from Seoul National University's Department of Computer Science and Engineering set out four levels instead of a single rule for the entire department. The instructor chooses a level based on the skills the assignment is intended to assess[1].
| Level | Permitted uses | Disclosure required |
|---|---|---|
| Level 0 Prohibited use | None | Not applicable |
| Level 1 Assistive use | Grammar correction, translation, wording improvements, and interpretation of error messages | Yes |
| Level 2 Use as a tool | Writing code, debugging, and research | Yes |
| Level 3 Unrestricted use | No restrictions | Yes |
These levels apply only to courses in the Department of Computer Science and Engineering
The guidelines state that, even within the same course, Level 1 or Level 2 may be assigned to work that does not assess implementation skills, such as research or reports. These guidelines apply only to undergraduate and graduate courses offered by the Department of Computer Science and Engineering. They do not cover theses or dissertations. The second semester of the 2026 academic year is a period for introducing and piloting the guidelines, with formal implementation beginning in the first semester of the 2027 academic year[1].
The names and number of levels also vary by university. Myongji University requires each assignment to specify one of three levels: prohibited use, conditionally permitted use, or encouraged use[7]. Hallym University's guidelines for instructors set out five categories, ranging from fully permitted to fully prohibited[8]. It is a mistake to apply another university's scale directly to your course.
Where to look for the rules for your assignment
Start with the most specific documents and work toward the broader ones. The guidelines from SNU's Department of Computer Science and Engineering also state that the criteria adjusted by the instructor and set out in the syllabus take priority[1].
- 1. Start with the assignment instructions. Look for AI provisions in announcements, assignment descriptions, and grading rubrics.
- 2. Check the AI section of the syllabus. SNU's Department of Computer Science and Engineering requires the level to be specified under "Policy on the Use of Generative AI Tools"[1].
- 3. Look for department or college guidelines.
- 4. Read the university guidelines. Check how violations are handled and what principles govern disclosure.
- 5. If you still cannot find the rules, ask your instructor. Korea University's guidelines say that, when the rules are unclear, students should check with their instructor in advance rather than decide on their own[4]. Including three details makes it easier to get an answer: the assignment name, the tool you want to use, and where you want to use it. Asking by email leaves a record of the answer.
What to include in your submission
If the course specifies a format, follow it first. If there is no set format, use the items that commonly appear in the published guidelines. SNU's Department of Computer Science and Engineering requires students to list the tools used, where they were used, and how the results were verified at the end of the submission[1]. Korea University requires tool names, when they were used, and the scope of use, with a clear distinction between the student's own work and AI's contribution[4]. Yonsei University provides examples of source disclosure that include the model, prompt, and platform[5].
Four common misconceptions
Thinking that disclosure will cost you marks. The guidelines from SNU's Department of Computer Science and Engineering state that using AI within the permitted scope and disclosing it will not, in itself, count against you in assessment. Failing to disclose its use is a violation. An omission alone is treated as a procedural violation rather than academic misconduct, but repeated omissions or clear intent to conceal are handled under academic regulations[1]. We could not confirm the same statement in other universities' documents.
Thinking that AI grammar correction does not need to be disclosed. A research-related footnote in SNU's guidelines states that, if AI was used only to correct foreign-language expression or check grammar, failure to disclose that use does not by itself constitute research misconduct such as plagiarism[2]. Course rules may differ. The Department of Computer Science and Engineering's Level 1 requires disclosure even for grammar correction[1].
Thinking that all you need to do is avoid detection. Korea University's guidelines say that AI detection tools should be used only as reference material, not as the sole basis for a judgment[4]. There are other ways to check, too. SNU's Department of Computer Science and Engineering may use oral checks for assignments at Level 2 or above, and an inability to explain your submission can be used as evidence[1].
Thinking that you can enter assignment questions into AI tools. Korea University's guidelines prohibit entering nonpublic learning materials and assessment questions into external AI tools[4]. This is because the content may be stored or reused.
References
- [1] Seoul National University's Department of Computer Science and Engineering, "AI Use Guidelines for Seoul National University's Department of Computer Science and Engineering", October 7, 2026. (In Korean.)
- [2] Seoul National University, full text of the adopted "Seoul National University AI Guidelines", established on January 1, 2026. (In Korean.)
- [3] Korea University, "Basic Guidelines for Using AI Such as ChatGPT", March 15, 2023. (In Korean.)
- [4] Korea University Undergraduate College Remote Education Center, "Guidelines for Using AI", March 30, 2026. (In Korean.)
- [5] Yonsei University, "Guidelines for Using Generative AI ver 1.0", May 2024. (In Korean.)
- [6] University of Seoul Office of Academic Affairs, "University of Seoul Guidelines for Using AI", May 2026. (In Korean.)
- [7] Myongji University, "Guidelines for Using Generative AI", established in August 2026. (In Korean.)
- [8] Hallym University Center for Education Innovation, "Guidelines for Instructors on Using AI in Courses". (In Korean.)
Frequently asked questions
My syllabus says nothing about AI. Can I use it?
It is safer to ask your instructor than to assume AI is permitted. That is what Korea University's guidelines advise[4]. SNU's Department of Computer Science and Engineering states that AI use will not be called into question later if the instructor has not specified a level, but that rule applies to that department[1].
Should I put my AI use disclosure in the references?
It depends on the university. SNU's Department of Computer Science and Engineering requires a separate disclosure at the end of the submission[1], while Yonsei University provides examples of source disclosure[5]. If your course does not specify a format, check with your instructor.
Do the same rules apply to theses or dissertations?
The guidelines from SNU's Department of Computer Science and Engineering exclude theses or dissertations and qualifying examinations, which are governed by the university guidelines and research integrity regulations[1].
Can I delete my drafts and AI chat history?
It is safer to retain records of your process as well. SNU's guidelines recommend keeping records of the assignment process, drafts, and revisions[2]. Myongji University's guidelines for instructors include a method of assessing contributions through shared links to AI chat logs submitted by students[7].