Teaching & Learning
- Grinnell does not impose a single college-wide rule banning AI in graded coursework
- University policy requires students to cite or acknowledge generative AI as a source, while faculty are free to set course-specific rules that may prohibit AI entirely, limit it to certain uses, or intentionally incorporate it into assignments
Grinnell’s Academic Honesty policy does not prohibit the use of generative AI by students, but requires that it be cited/acknowledged as any other source would be.
Faculty are free to develop their own course policies regarding the use of AI by students in their class. You may choose to prohibit the use of AI for any use or for only specific uses. You may choose to have students engage with generative AI in purposeful and direct ways for assignments, this is up to you and what you deem appropriate for your teaching.
You should also state when and how students may (or may not) use generative AI/LLMs/chatbots as a resource (e.g., is it allowable for writing a draft? Generating ideas? “Polishing” or condensing writing? Translation?).
- For exams and assessments, Grinnell’s baseline academic honesty rule is that tests are closed-book and assigned work is independent unless the instructor says otherwise
- Guidance to faculty specifically recommends stating the AI policy for each assignment or assessment, so AI use during exams is governed by instructor-defined course rules plus the college’s general citation expectations
Tests or examinations are closed-book unless the instructor states otherwise.
Any assigned work is to be done independently unless the instructor states otherwise.
Best practice is to state your specific AI policy for each assignment or assessment in your class.
- The university does not set a uniform student-wide rule specifically for AI as non-graded study help
- Instead, faculty are expected to tell students when and how AI tools may or may not be used as a resource, and teaching guidance explicitly treats uses such as generating ideas or translation as matters for course policy
You should also state when and how students may (or may not) use generative AI/LLMs/chatbots as a resource (e.g., is it allowable for writing a draft? Generating ideas? “Polishing” or condensing writing? Translation?).
Faculty are free to develop their own course policies regarding the use of AI by students in their class.
- For programming work, Grinnell applies its general academic honesty expectations rather than a separate college-wide coding rule
- Students must cite AI use, and instructors may decide whether AI is allowed or prohibited in particular courses or assignments, including computer programs submitted for grades
When you submit a piece of work (whether a paper or paper draft, report, examination, homework, computer program, creative project, or other assignment) for a grade, you are claiming that its form and content represent your own original work produced for this assignment, except where you have clearly and specifically cited other sources.
Grinnell’s Academic Honesty policy does not prohibit the use of generative AI by students, but requires that it be cited/acknowledged as any other source would be.
Faculty are free to develop their own course policies regarding the use of AI by students in their class. You may choose to prohibit the use of AI for any use or for only specific uses.
Research
- For grant proposal preparation, Grinnell does not set a final standalone internal AI rule but directs applicants to follow funder requirements because norms are changing
- The handbook states that NSF encouraged applicants to disclose whether and how generative AI was used in proposal development, and it notes that NSF and NIH ban generative AI in proposal review
At the time of this writing (January 2024), funding agencies and foundations are just beginning to issue guidance on the use of generative artificial intelligence (AI) in proposal preparation.
In December 2023, the National Science Foundation (NSF) issued a media release encouraging applicants to indicate in the project description the extent to which, if any, generative AI technology was used and how it was used to develop their proposal.
Both the NSF and the National Institutes of Health have explicitly banned the use of generative AI in the proposal review process.
While future editions of this handbook will provide updated guidance on the use of generative AI in proposal preparation, applicants should be advised that the norms around generative AI use are evolving rapidly and they should be attentive to any guidance or regulations offered by funders on this matter.
Academic Integrity
- Grinnell requires disclosure of AI use through citation or acknowledgment
- The academic honesty materials treat generative AI like any other source, and faculty are advised to specify citation expectations and preferred citation systems in their courses
Grinnell’s Academic Honesty policy does not prohibit the use of generative AI by students, but requires that it be cited/acknowledged as any other source would be.
Academic honesty is fundamentally about students accurately representing the work they submit as their own and appropriately citing or acknowledging the use of other resources or contributors to that work. Grinnell’s Academic Honesty policy sets basic assumptions and expectations regarding collaboration, outside resources on exams, and citation/acknowledgement, including for LLMs and other generative artificial intelligence (AI).
Indicate your preferred systems for citation.
- Grinnell’s AI working materials also indicate that policy work is being developed around academic integrity, including AI detection
- When there is evidence of uncited AI use, instructors should refer the case to the Committee on Academic Standing rather than decide the matter themselves
- Faculty are told not to impose their own penalties; cases are adjudicated through the college process, and instructors then apply any required grade outcome
If you have evidence of uncited AI use by a student, you should submit the case to the Committee on Academic Standing.
o Do not make your own determination or impose your own penalty for an academic honesty violation. It is important that students are treated consistently and according to our processes.
o Cases are typically adjudicated within two weeks of submission. You will be notified by email of the decision (responsible or not responsible for a violation of the academic honesty policy) and any outcome that you are required to apply to the student’s grade as a result of a responsible finding
Formulating recommendations for policy/guidelines, including those related to academic integrity (including AI detection) and faculty and teaching staff use of AI in their own work (grading/feedback, assignment creation, etc.
Institutional & Administrative
- Faculty have substantial autonomy in deciding how to use or respond to AI in teaching, including assignments, grading methods, and course policies
- Institutional planning also explicitly includes future guidance on faculty and teaching staff use of AI in grading, feedback, and assignment creation
- For employees using approved AI tools in administrative work, the college requires human oversight: staff remain responsible for reviewing content, checking accuracy, following approvals, and exercising professional judgment
The Center for Teaching and Learning works to support instructors at Grinnell in the many ways they may decide to approach AI tools in their courses: changing assignments or grading methods in response to AI, formulating class policies about AI, assigning readings or activities in the mode of Critical AI studies, working with students to understand the new complementarities between human and computational work, and more.
Grinnell’s instructors have a good deal of autonomy in making these decisions, either individually or collectively, as in the example of the Department of Art History’s AI Policy.
Copilot can support your work, but employees remain responsible for reviewing content, verifying accuracy, following approval processes, and exercising professional judgement.
Formulating recommendations for policy/guidelines, including those related to academic integrity (including AI detection) and faculty and teaching staff use of AI in their own work (grading/feedback, assignment creation, etc.
- The stated basis is security, privacy, compliance, and alignment with institutional data governance expectations
- Grinnell designates Microsoft 365 Copilot as the only approved and institutionally secure AI tool for work involving college data
- Restricted data, including information identifiable to a student, must not be shared with AI tools outside the college’s Copilot environment
For data security and privacy, Microsoft 365 Copilot is the only institutionally secure and approved AI tool for work involving college data.
Public or consumer AI tools are not secure and therefore not approved for administrative work involving College data.
One important restriction on our use of AI tools is that restricted data, including information identifiable as connected to an individual student, should not be shared with any AI tool outside of the Copilot that is part of the College’s Microsoft 365 environment, as explained in the College’s page on AI in the Grinnell Context.
Copilot aligns with institutional data governance expectations.
- Its current strategy work also includes exploring tools, making adoption recommendations, supporting faculty development, and developing policy guidance
- Grinnell has an institutional AI governance effort centered on an AI Working Group and participation in AAC&U’s AI, Pedagogy and the Curriculum institute
- The college states that the working group is tasked with identifying uses and concerns, articulating ethical and practical issues, creating institutional principles, and suggesting a governance structure
The AI Working Group’s membership includes representative faculty from all three academic divisions and Executive Council, as well as staff from Senior Leadership, Information Technology Services, Academic Affairs, Communications and Marketing, Accessibility and Disability Services, and Staff Council.
Identify and document areas of interest and concern regarding the uses of AI at Grinnell College, whether by students, faculty, or staff.
Identify and articulate the practical and ethical issues pertaining to AI use at Grinnell.
Formulate a set of institutional principles to guide the use of AI in accordance with Grinnell College’s mission and values.
Suggest a governance structure to oversee the development and evolution of AI use at the College.
As part of this effort, the team is:
Creating inventories of faculty AI use in their pedagogy and teaching work
Exploring AI tools and making recommendations for adoption
Developing institutionally supported mechanisms for faculty development and discussion around AI in the classroom, including AI literacy, basic student AI competencies, the use of AI in teaching writing, critical thinking, and accessibility.
Formulating recommendations for policy/guidelines, including those related to academic integrity (including AI detection) and faculty and teaching staff use of AI in their own work (grading/feedback, assignment creation, etc.
DocuMark: Responsible AI Use for Academic Integrity
Knowing your institution's AI policy is step one. DocuMark helps enforce it fairly by empowering universities to manage AI-generated content, prevent cheating, and support student writing through responsible AI use.
Common Questions About Grinnell College's AI Policies
Grinnell College has defined AI policies in 10 of 12 categories, with an overall coverage score of 83%.
Grinnell requires disclosure of AI use through citation or acknowledgment. The academic honesty materials treat generative AI like any other source, and faculty are advised to specify citation expectations and preferred citation systems in their courses.
When there is evidence of uncited AI use, instructors should refer the case to the Committee on Academic Standing rather than decide the matter themselves. Faculty are told not to impose their own penalties; cases are adjudicated through the college process, and instructors then apply any required grade outcome. Grinnell’s AI working materials also indicate that policy work is being developed around academic integrity, including AI detection.
Grinnell designates Microsoft 365 Copilot as the only approved and institutionally secure AI tool for work involving college data. Restricted data, including information identifiable to a student, must not be shared with AI tools outside the college’s Copilot environment. The stated basis is security, privacy, compliance, and alignment with institutional data governance expectations.