College of Arts and Sciences Artificial Intelligence Policy

PURPOSE AND VALUES

As a liberal arts institution, one of The University of the South’s main objectives is to instill in our students the abilities to problem solve and to think critically while upholding high standards of honor, trust, and ethical responsibility. Courses at Sewanee are intentionally designed by instructors to support this broader mission by training students in discipline-specific knowledge and skills. These learning experiences often require varied levels of challenge, practice, and exposure curated by expert faculty to foster student growth. Advancing technologies have the potential to undermine those skills if they are not used responsibly. The purpose of this policy is to aid coordination in our instructional design in light of particular advancing technologies to ensure the holistic education of students is not compromised.  

“Artificial Intelligence” is an umbrella term that refers to automated systems which are designed to mimic human thinking or outputs. Generative Artificial Intelligence is a subtype of such systems that use many layers of processing and complex algorithms in order to produce new content, such as writing, images, or code, thereby substituting for work that would otherwise require a significant amount of personal and original effort. In this document, “AI” specifically refers to Generative AI tools. 

The purpose of this policy is not to take a wholesale ethical stance in favor of or against AI, but instead to support an intellectual environment that is conducive to responsible and respectful dialogue of the substantive and ethical issues involved. The College recognizes that AI may present opportunities for students and faculty to benefit from reflectively engaging with these technologies as part of their intellectual endeavors. For example, AI tools may help individuals develop the capacity to apply expert judgment to new content. AI tools may allow for a new level of research responsibility. AI tools can also aid breadth of inquiry. The College also recognizes that many features of AI complicate the straightforward incorporation of its use in the classroom. To name a few, information is taken from existing work, often without the explicit permission of the authors, clouding issues of fair use. AI information can be inaccurate, outdated, and biased. Many resources are also required to enable the large amount of processing, including electricity and clean water. These resources can tax our existing structures that were not built or updated for the new demand. Finally, there has not yet been an opportunity for long-term studies on the effects of humans’ reliance on these rapidly developing technologies.  Due to the complexity of these considerations, the College will establish a fellowship program, the Fellows for AI Readiness, aimed at collecting information, communicating guidelines, sharing resources, fostering dialogue, and updating policies in support of our intellectual community. 

The College standard is that the use of AI at Sewanee should never circumvent or undermine the conditions that make for an excellent liberal arts education: an education that engages both with the enduring questions of humanity and the needs of our current society through honor, stewardship, and citizenship. 

SCOPE

The policy applies to all faculty, instructors, and students in the College of Arts and Sciences, with some policies having a smaller scope as indicated below. 

Given the complexity of the factors involved in AI use and the speed with which advancements occur, the Fellows for AI Readiness (FAIR) will lead the college in an annual review of this policy.

Definitions

general policy

All faculty, instructors, and students in the College of Arts and Sciences must align their use of AI technology with Sewanee’s holistic liberal arts education; each use must support development in critical thinking, honor, stewardship, and citizenship. 

Faculty and instructors must follow the integrity standards of their discipline or industry. 

Student use of AI for any coursework is prohibited unless explicitly permitted by the instructor.  Instructors are responsible for determining whether, how, and to what extent AI may be used in each course, assignment, or learning activity.  

Instructor and student uses of AI in a course must align with course learning goals and be cited according to discipline or industry standards. This will typically include indicating the extent of AI reliance and the particular resource(s) used. Users are expected to familiarize themselves with the shortcomings of AI and to follow precautions (e.g., fact-checking). 

All members of the Sewanee community are responsible for safeguarding privacy and protecting intellectual property. A common practice of AI tools is that any content related to the use of the tool may be stored, reviewed, or used to train future models. This includes content generated by the tool in response to the users’ queries, and it includes content contributed by the user through direct input, uploading, or granting access permissions. In some cases, this information may also be accessible to other parties. AI tools should not be considered confidential or secure environments. Students and faculty should avoid entering sensitive, confidential, proprietary, or personally identifiable information into AI platforms. No use of AI should conflict with Sewanee’s Privacy Policy or Acceptable Use Policy.

Instructor policy

Each instructor must model honorable engagement with AI according to the highest integrity standards of their discipline or industry. To do so, instructors will: 

1. Disclose any use of AI in their research with citations according to the discipline standards set forth by their intellectual community, e.g., journal submission requirements.
2. AI should not be used to assess students. Instructors will maintain the pedagogical integrity of a small liberal arts college by providing feedback and grades that reflect their own expert evaluation. 
3. Communicate the AI policy for their course in the Syllabus pertaining both to student and instructor use, and explicitly tie that policy to the course’s learning objectives. Professors should be careful to follow AAUP guidelines regarding academic freedom as outlined in the Personnel Procedures. The kinds (e.g., chatbots, image generators) and models of AI (e.g., Claude 4.6, DALL-E 3) allowed must be specified.

4. Given the rapidly evolving nature of technological tools, this policy cannot provide an exhaustive list of permitted and prohibited applications. Accordingly, the default position is that the use of AI tools is not permitted unless expressly authorized by the professor for that assignment. If AI is explicitly permitted, then: 

   a. The AI use policy must be reiterated on each assignment description to ensure clarity of requirements to students. Again, the policy must include the kinds and particular models permitted, and should indicate the scope of use as well.

   b. Guidelines on how to cite AI should be provided to the student. 

   c. Where uncertainty exists regarding a specific tool, students are responsible for seeking clarification and approval from the professor prior to its use, including confirmation of the scope of any permitted use. Professors should view these discussions as mentorship opportunities and should seek to facilitate these conversations as appropriate. 

   d. Ensure equitable use of AI by comparing student needs to College or public resources and requesting any necessary access.  

   i. When AI is required, the professor should work with Student Accessibility Services to ensure equal opportunities amongst all students regardless of means or background.

   ii. When AI use is optional, students who opt out should be provided an alternative resource or support to create equity.

5. When AI use is banned, instructors may take measures to ensure an appropriate level of accountability insofar as those measures are conducive to equity.
6. In holding students accountable, instructors must make decisions on the basis of clear and convincing evidence. AI detectors are currently flawed and should not be used as the sole source of evidence. Instead, faculty should make efforts to familiarize themselves with their students’ work and discuss suspected violations with the student.

STUDENT POLICY

AI use is prohibited unless explicitly indicated by the instructor in the course syllabus or on an assignment. Each student must uphold the Honor Code, which includes completing their work as assigned and not obfuscating their work’s merit. Students will: 

1. Know and follow the AI policy for each class as indicated by the instructor.  
2. Reach out to the instructor before completion of assignment(s) if you have any doubt about any aspect of an AI policy. These conversations should be viewed as mentorship opportunities that are an essential part of a liberal arts education.
   a. Participate in a dialogue with the instructor where appropriate to ensure class policies are a good fit with the emerging realities of technology.
   b. Consult with instructors regularly regarding your AI use patterns to better train your judgment regarding the use of AI. 
3. When AI use is permitted on an assignment, follow the Student AI Guidelines in the responsible evaluation of each use of AI. Consider the learning objectives of the course, the ethical, environmental, and deskilling costs, as well as what you stand to gain from the AI use.  
4. Disclose and justify to faculty any use of AI according to the citation guidelines provided in the instructor’s policy.

CONSEQUENCES FOR INAPPROPRIATE USE OF AI

The goal of the policy is to preserve the integrity of Sewanee’s intellectual community and the rigor of a Sewanee liberal arts education. Nevertheless, the landscape of technology-assisted intellectual endeavors is quickly changing. This means that administration, faculty, and students need to prioritize communication and co-learning in their interactions. The following penalties are to be used only when reasonably informed individuals have acted inappropriately, not when a miscommunication or an unanticipated technological shift has caused an error.

RESEARCH MISCONDUCT

In cases where AI is inappropriately used in research, the procedures set out in the Research Misconduct Policy should be followed.

pedagogical misconduct

In all cases of pedagogical misconduct, students can follow the standard grievance procedure as outlined in Sewanee’s Student grievance decision tree. More particulars are outlined below for convenience. 

In cases where an instructor uses AI in a manner that undermines education in critical thinking, stewardship, honor, and citizenship, students should discuss with the instructor the ways in which they think the AI use is affecting their ability to learn. If they are unable to resolve the matter with the instructor, they may discuss the matter with the chair of the department.  Department chairs may discuss individually with the instructor or with the department as a whole as appropriate. If desired, the Fellows for AI Readiness may assist in this process. Finally, if the matter is still not resolved, then the student may contact the Office of the Dean of the College, who will collect information, at deancas@sewanee.edu. Disciplinary procedures will follow those outlined in Sewanee’s Personnel Procedures.  

If the student believes an assignment was incorrectly categorized as a misuse of AI and therefore received a lower grade, they should first discuss the matter with the instructor directly. If they are unable to resolve the matter with the instructor, they may discuss the matter with the chair of the department. If matters are still not resolved, they may submit a grade appeal for their final course grade.

ACADEMIC MISCONDUCT

Failure to meet the standards outlined in this policy constitutes inappropriate use of AI. The scope and severity of consequences will be determined at the instructor’s discretion and may include, but are not limited to: 

• Required revision or resubmission 
• Oral examination or reflective discussion 
• Failure of the assessment 
• Reporting to the Honor Council 

All incidents of AI misuse will be documented with the Honor Council as outlined below, regardless of whether they are escalated to an Honor Council proceeding.

Involvement of Honor Council

Inappropriate use of AI is a violation of the University Honor Code and may result in additional institutional sanctions, up to and including suspension from the University. While the instructor has discretion to engage in teachable moments with the student, the instructor should not lose sight of Sewanee’s institutional standards of honor and due process.  

If the student believes the assignment was incorrectly categorized as AI and received a lower course grade, they may submit a grade appeal. Otherwise, certain cases also can merit advancement to a submission to the Honor Council: 

• If the instructor thinks the violation merits failure of the course and possible suspension.
• If the instructor decides the violation warrants a failing grade on an assignment weighted as such to determine the student's passing of the course, the violation can be sent by the instructor to the Honor Council for additional deliberation. 
• If the student does not admit responsibility to using AI inappropriately, the professor can refer the case to the Honor Council to ensure due process.

Documentation Process

In accordance with the Honor Council’s established procedures, within 120 hours of discovery of any suspected AI use violation, professors must submit documentation with the Secretary of the Honor Council via a Documentation of AI Misuse Form. After submission, faculty have another 120 hours to update the case as escalated, resolved, or retracted; otherwise the case defaults to a formal Honor Council proceeding. If the College is not in session at the time of the detection, these one hundred and twenty (120) hour periods are to be measured from the resumption of the College session. It is recommended that during this second 120 hour timeframe the professor communicates with the student to better determine the scope and severity of the AI misuse and whether a teachable moment is appropriate.  

• Updating the form with ‘escalated’ amounts to a request for a formal proceeding.  
• Updating the form with ‘resolved’ amounts to a request for the Honor Council to keep the record. Further information about the assigned penalty must be included.
• Updating the form with ‘retracted’ amounts to a request for the Honor Council to delete the record. This is acknowledgement that the initial suspicion was mistaken. 

The purpose of this documentation is to ensure “teachable moments” are effective and to avoid multiple occurrences of AI misuse. If there are multiple offenses by a student, the Honor Council may initiate a formal proceeding.

RESPONSIBILITIES

The Fellows for AI Readiness (FAIR) will review the workings of this policy during the 2026-2027 Academic Year. The policy may be updated yearly according to their findings and the changing technologies and submitted to CAPC for endorsement.