Guidance Documents to Support Graduate Education

The Use of Generative AI in Graduate Committee-Based Exams, Projects, Theses, and Dissertations
Graduate School, The University of Alabama, December 2025

As the culminating experience of master’s and doctoral degrees, theses and dissertations assess multiple dimensions of a student’s academic skills, including but not limited to: formulating an original research question; conducting broad, thorough, accurate, and advanced research; advancing and testing a hypothesis or theoretical framework; assessing and analyzing results or conclusions; creating an original expression of art/music/imagination; creating multi-media products; or writing the study at the expected level of sophistication, professional fluency, and precision expected of scholars possessing a graduate degree.

In degree programs that require a committee-based project, thesis, and/or dissertation, students are expected to demonstrate the above skills through both a written document and an oral defense. Regardless of the process used to compose the document, students are responsible for the full content of the document and are expected to demonstrate, in the oral defense, their ability to engage with the concepts and evidence provided in the document. Because the development and mastery of advanced academic skills is expected not only during the degree program but throughout the scholar’s subsequent professional career, the primary responsibility of authorship resides with the student. For these reasons, over-reliance (as defined by the academic discipline and the program or advisory committee faculty) on generative AI is strongly discouraged.

Generative AI refers to artificial intelligence systems that can create new content, including text, images, audio, video, code, or data outputs, in response to user prompts or inputs. This includes, but is not limited to:

• Large language models (e.g., ChatGPT, Claude, Gemini, Copilot)
• AI writing assistants that generate substantive content beyond basic grammar/spelling correction (e.g., NotebookLM)
• AI coding assistants (e.g., GitHub Copilot, CodeWhisperer)
• AI image generators (e.g., DALL-E, Sora, Veo 3, Veo 3.1, Midjourney, Stable Diffusion)
• AI audio and video generation tools (e.g., Sora, Veo 3, Veo 3.1)
• AI-powered data analysis tools that generate interpretations or conclusions

Different academic disciplines will have distinct expectations regarding the level to which the exam/thesis/dissertation process can or should be aided by generative AI technologies. Some examples of generative-AI-integrated work might include:

• Initial brainstorming
• Initial literature review
• Development of theoretical framework
• Design of research methodology
• Data analysis
• Data transcription and coding
• Generation of static graphics (charts, images, and diagrams) or dynamic graphics (animations, interactive visualizations, or simulations) or video
• Generation of audio content
• Generation of software code
• Translation
• Proofreading/editing/revision for improvements in style, organization, clarity
• Proofreading/editing/revision for accuracy and sophistication
• Generating full paragraphs/pages/chapters from the prompts or data provided

A thorough disclosure of how generative AI was used in the exam/project/thesis/dissertation process is required of all students who submit generative-AI-integrated work.

As with all matters of academic integrity, students are expected to adhere to the guidelines of their degree program. Faculty members should remain attentive to work that is technically inconsistent or unusual, that may signal inappropriate use of generative AI. Unauthorized use of generative AI tools for scholarly work may be considered an offense under UA’s Academic Misconduct Policy and/or Research Misconduct Policy.

All users need to be aware of the potential inaccuracies, fabrication of sources and citations, and biases of generative AI tools since they are trained on data from the public internet.

Departments may adopt a shared policy on the use of generative AI for committee-based exams (including qualifying exams, comprehensive exams, and candidacy exams), projects, theses, and dissertations. However, many decisions about the appropriate use of generative AI are topic-dependent. Thus, variation from a department-wide policy is permissible, but should be documented so that the guidance to students is clear and shared among the members of the faculty committee. The guidance to students should be provided as early as possible in the student’s research or exam-preparation work.

Students should be made aware that current copyright regulations around generative-AI-produced content are unclear. Copyright violations might also occur when AI supplies text from unattributed sources. Additionally, future detection tools might point towards generative-AI-sourced text that was inaccurately documented by the student author. Limitations regarding future publication may arise. Thus, caution is urged for all students when using generative AI in the composition of their manuscripts.

Examples of statements that programs might use/adapt include:

• The use of generative AI is permitted for brainstorming, initial bibliographic searches, and the production of graphics (charts, images, etc.). The student is responsible for verifying all bibliographic information, and for any omissions; this responsibility cannot be transferred to generative AI. Students are also responsible for ensuring that information from generative AI has not been plagiarized from some other source, given that generative AI often reproduces text from prior sources. Students are responsible for writing the exam/project/thesis/dissertation manuscript themselves, although generative AI may be used for editing. Entire paragraphs written by generative AI – even if from student-submitted prompts – are not allowed, since students (with faculty mentorship) are expected to develop and refine their own independent scholarly voice and style. Students are expected to include a Generative AI Disclosure Statement in the opening pages of the manuscript, detailing how generative AI supported the creation and writing of the exam/thesis/dissertation.

• The use of generative AI is permitted for data analysis and computation, as well as for research. However, students are responsible for the accuracy and completeness of all information generated by AI. Students are responsible for fact-checking any AI-generated content.

• The use of generative AI is permitted for developing/writing/debugging/proofreading code; however, responsible AI-assisted coding is itself a graduate learning objective. Students are expected to integrate AI tools thoughtfully and transparently, using them to deepen technical understanding rather than bypass learning, and to demonstrate maturity in deciding when AI assistance is appropriate versus when independent implementation is required. Effective use includes employing AI to explore design options and prototype solutions while maintaining mastery of foundational programming principles; critically evaluating and refining AI-generated code for correctness, efficiency, security, ethics, reproducibility, and maintainability; and ensuring clarity and documentation through human-readable logic, comments, and version control. As part of their development, students should operate as AI-informed learners who use AI to expand exploration, as evaluators who rigorously assess and improve AI outputs, and as innovators who refine workflows and contribute creative or methodological improvements to AI-enabled programming practices. Students must be able to explain and justify their coding decisions, document AI assistance, and demonstrate full conceptual command and independent authorship. Ultimately, AI may enhance rigor, creativity, and capability, but it cannot replace disciplinary expertise, academic integrity, or critical thinking; students remain fully responsible for the validity, originality, and functionality of all code submitted.

• The use of generative AI is not permitted at any phase of the exam/project/thesis/dissertation work. All research and writing must be produced by the students themselves. Use of generative AI tools such as ChatGPT in the research, writing, or revision of an exam/thesis/dissertation manuscript constitutes academic misconduct.

All students who use generative AI during the exam/project/thesis/dissertation process must include a disclosure statement, the exact nature of which should be agreed upon between the student and the faculty at the start of the project. A sample structure for such a disclosure statement might be:

During the preparation of this exam/project/thesis/dissertation, the author has utilized [Generative AI Tool Name and version/date], a language model created by [Generative AI Tool Provider]. The [Generative AI Tool Name] was used for purposes such as [e.g., brainstorming, grammatical correction, writing paraphrasing, citation, etc.].”

Students should be made aware that most professional publication outlets will also require such a statement. Thus, the accuracy of these disclosures, documenting from the outset the ways in which generative AI was used, is of the utmost importance. As an example: the Journal of Information Systems editorial policy contains the following statement: “Authors need to disclose the use of generative AI and AI-assisted tools in their work. Use of AI and AI-assisted writing tools must be consistent with the AAA policies on Authorship and Plagiarism, as well as other requirements listed in the AAA’s Publications Ethics for Academic Research.”

Data Privacy and Confidentiality: Students must ensure that any data entered into generative AI tools complies with UA policies including IRB protocols, data use and software licensing agreements, intellectual property regulations, and confidentiality requirements. Sensitive, proprietary, or identifiable information should not be uploaded to commercial AI platforms without appropriate safeguards. Usage must also comply with UA’s Office of Information Technology list of approved and prohibited AI tools and the Office of Research & Economic Development policies regarding intellectual property as well as data governance policies regarding quality, privacy and confidentiality.

Research Contributions to AI Models and Datasets: While this guidance focuses on the permissible use of generative AI in exams/projects/theses/dissertations, there may be graduate-level research that contributes new knowledge regarding the models and datasets that underpin AI systems. Thus, the current guidance should not be interpreted as discouraging student projects that advance the research, development, or application of generative AI, whether for purposes of discovery, improvement, or comparative evaluation of AI-driven methods.

Training and Resources: The UA Teaching Academy (UATA) provides training resources on ethical use of generative AI in teaching and learning, as well as verification techniques and best practices for disclosure. Faculty and graduate students are encouraged to participate in these training opportunities.

Document source: Ad Hoc Committee on Generative AI Use in Graduate Committee-Based Exams, Projects, Theses, and Dissertations, Fall 2025.

---

FULL DOCUMENT TEXT:

Guidance regarding Publications and Public Professional Activities by UA Students and Postdoctoral Scholars

Graduate School, The University of Alabama, December 2025

University-level education engages students and postdoctoral scholars (postdocs), as emerging scholars, to deepen their disciplinary expertise, develop their own voices and creativity, and acquire the skills to produce new knowledge. The mentor-mentee relationship values the student’s independence, but also relies on the mentor’s professional experience and guidance, to ensure rigor and foster the student’s future career opportunities. Guidance from faculty mentors on the content and readiness of manuscripts or products generated by students to be shared in a public forum, as well as discussion of potential intellectual property concerns and publication outlets, protect the interests of the student, the faculty mentor, and The University of Alabama.

This guidance is intended to uphold research integrity, authorship ethics, and responsible publication practices consistent with UA policy. Poor publishing or dissemination decisions can harm a student’s professional prospects through potential retractions, allegations of plagiarism, or violation of grant or contract provisions. Predatory journals or venues pose financial and reputational risks to the student and can reflect unfavorably on the student’s home department and mentors. Because publications or other products listing The University of Alabama as an affiliation are publicly associated with the institution, students must ensure that such use is judicious and appropriate and does not imply University endorsement. Additionally, all students, faculty, and staff are responsible for compliance with UA’s Intellectual Property Policy, Responsible Conduct of Research standards, and sponsor-publication requirements. Strategic, well-advised choices about what, where, and when to publish or present one’s work can position students advantageously to compete for jobs and opportunities in the future.

The following guidance is intended to help foster conversations around the publication of manuscripts, or other sharing of research, scholarly projects, or creative activity, as a UA student, with appropriate checks and balances between mentor and mentee to safeguard the academic reputations of all parties. Nothing in this guidance is intended to limit students’ academic freedom or their right to pursue independent scholarly or creative work, consistent with applicable University policies.

REQUIREMENTS

• If the research involves substantial use of UA resources, appropriate faculty or administrative oversight and approval is required before submitting any related research or activity for presentation or publication. “Substantial use” generally refers to use of resources beyond what is normally provided to students.  Use of resources such as staff support, collaboration with other members of the University community in their professional capacities, specialized equipment or space, or University funds allocated to the project constitutes “Substantial Use”. Works developed with no more than incidental or insignificant use of University resources (as defined in the University’s Intellectual Property Policy) are not considered University-Assisted Works and remain the property of the student author(s).  Policies regarding ownership of intellectual property are also outlined in UA’s Intellectual Property Policy. These policies apply to the use of UA facilities and to collaboration with UA employees.
• Students may identify their affiliation as “The University of Alabama” in scholarly or creative works only when the work was substantially conducted under UA supervision or with UA resources. Use of the UA name, logo, or marks must comply with the University’s Brand Standards as maintained by Strategic Communications and may not imply University review or endorsement of the work or its conclusions.
• Students supported by University or externally sponsored funds must comply with sponsor and University publication policies, including data ownership and confidentiality obligations. Student employees typically do not own data generated in the course of sponsored research and must obtain PI or faculty approval in advance of any publication that cites or uses this data.
• Faculty members’ names may not be listed as co-authors or in acknowledgments without their explicit permission and the opportunity to review the final manuscript. Authorship should reflect genuine intellectual contribution consistent with national authorship standards. (See, for example, resources provided by the Declaration on Research Assessment, or the International Committee of Medical Journal Editors.)

GUIDELINES

In situations not covered by the above requirements, UA students are nonetheless strongly advised to discuss publication opportunities and plans, or other public professional activities, with their mentors. Mentors or departments may require this consultation as a condition of their assistantship support. Regarding publications, “consultation” includes discussing the manuscript’s content, its readiness for publication, and/or the choice of publication outlets. Departments may require students to vet the journal or publisher through credible resources such as those provided by UA Libraries (e.g., Cabell’s Directory of Publishing Opportunities). As noted above, publishing in deceptive or non-peer-reviewed outlets can harm both the student’s and the University’s reputations.

Similar consultation is encouraged (and may be required by the department) when students apply to present their research or scholarship at professional conferences, or to present their work at exhibits and other professional/creative venues, using the UA name and/or affiliation. This is even more strongly recommended (and may be required) when UA resources support students’ participation in the activity. Such conversations provide mentorship not only in the research area but also in professional ethics and integrity.

UA students who are neither employed by the university nor utilizing university resources may have ownership over the manuscripts or products they generate; however, they may only list their affiliation with the University of Alabama on those products in accordance with UA’s policy on Publication, Authorship, and Institutional Affiliation (currently in final approval stage with the Division for Research). In these cases, students are strongly advised to avail themselves of faculty mentorship when deciding what, where, and when to present and/or publish their work.

Students should be aware that, in most cases, they will be required to transfer their copyright to the publisher in exchange for publication. Faculty advisors or UA Libraries can provide guidance about Open Access publication, which involves different copyright and license regulations.

Faculty members serving as research advisors may require consultation regarding publication or other public professional activities, as a condition of their mentorship. Students should ensure they are aware of their department’s policies regarding student presentations and publication.

Colleges or professional fields may have their own guidance in terms of authorship issues. These guidelines may address who should or must be listed as co-authors, the order of collaborators’ names, and other disciplinary norms. In the absence of specific guidelines, early and frequent conversations between mentors and mentees are encouraged to avoid subsequent conflicts. If conflicts arise, they should be discussed with department or college leadership, as outlined in UA’s policy on Publication, Authorship, and Institutional Affiliation.

An additional resource regarding authorship credit and ordering is the taxonomy endorsed by the American National Standards Institute and National Information Standards Organization (ANSI/NISO).

Questions or conflicts that arise regarding the use of the UA name and/or public professional activities, or questions about this guidance document, should be addressed through consultation with the department chair and/or the college dean’s office.

Students should be aware that, when listing The University of Alabama as their affiliation, they are representing the institution’s professional standards for research and/or creative activities. The inclusion of the following statement is encouraged when appropriate, as outlined in UA’s policy on Publication, Authorship, and Institutional Affiliation:

“The views expressed are those of the author(s) and do not represent the views of The University of Alabama.”

This document was generated by UA’s Ad Hoc Committee on Student Publishing, Fall 2025.

Professional Activities

Mentoring Compact Template

Mentorship compacts are a useful tool to enhance graduate student success, by making explicit the expectations related to the advisor-advisee relationship. Such documents can clarify how communications should be managed, how feedback is shared, what deadlines or milestones should be kept in mind, and expectations related to professionalism, among other topics. The Center for the Improvement of Mentored Experiences in Research, a nationally recognized leader in the area of graduate mentorship, maintains a Library of Mentorship Agreements, offering ample resources for constructing your own compact. In addition, the Graduate School has consulted with UA faculty to create a broad template (LINK/POPOUT) that may inform the development of your compact.

(Document forthcoming)

Incompletes (IN) Agreements

As outlined in the Graduate Catalog, faculty may choose to assign an Incomplete grade (IN) at the conclusion of the semester. As the policy notes, “All ‘IN’ grades must be removed by the conclusion of the subsequent regular semester” but instructors may specify a shorter time frame. Instructors are encouraged to utilize this Incomplete Agreement form [LINK/POPOUT], or a similar written agreement, in order to ensure that students understand the requirements for removal of the “IN” grade. For further information, please consult the full Incompletes policy in the Graduate Catalog. 

(Document forthcoming)

Cross-Listed 400/500 Level Courses

While academic programs are generally encouraged to offer distinct undergraduate and graduate courses, there are circumstances in which academic programs combine select graduate and undergraduate courses together as one course. Cross-listed 400/500-level undergraduate/graduate courses refer to the simultaneous offering of two courses, one undergraduate (400-level) and one graduate (500-level), that share a class experience but maintain distinct academic expectations.

To maximize the educational experience at both the graduate and undergraduate level, these cross-listed courses must adhere to particular guidelines, as outlined on the UA Registrar website. The UA Teaching Academy website amplifies these guidelines by offering strategies and examples for course construction, as well as teaching tips for instructors who are addressing these blended student groups.