Developing AI literacy is important and can be taught alongside our learning outcomes. There are many valid positions for the potential role of LLMs in the writing classroom, and all instructors will need to engage with the salient pedagogical and ethical issues involved to design a policy and practice aligned with their philosophies, course goals, and students’ needs.

Below we offer the following as resources for PWR instructors:

• A more restrictive AI/LLM use policy, for two reasons:
  • ○ First, students in 100-level composition courses need to develop their writing/composing abilities for themselves; we don’t want to make students’ composing processes dependent on AI/LLMs; and
  • ○ Second, in the short time between orientation and beginning to teach, we recognize that there are limits in what new-to-PWR instructors may be able to engage with around AI/LLMs in relation to their pedagogy. We anticipate that instructors will adapt and develop this language as they engage with the relevant issues and build policies that work for their course.
• Some initial guidance for you in determining how/if/when to incorporate AI/LLMs into their course
• Sample syllabi policy grouped by limited, nuanced use of AI/LLMs to use of AI/LLMs encouraged to other prohibitive policies

Note: If this is your first quarter teaching in PWR and/or your experience with AI/LLMs is limited, we encourage you to start with the “AI/LLM Restrictive Usage Default Policy,” provided below, while engaging with the broader pedagogical and ethical questions surrounding AI use to develop your own language and approach for use in subsequent quarters.

AI/LLM Restrictive Usage Policy Default

Large language models (LLMs) such as ChatGPT, Claude, and Gemini can assist in research and writing. They are also being increasingly deployed across other sectors of society, and we all have a stake in understanding how they work, what they are used for, and what benefits and harms they produce. As you develop your own writing practice, you may find ways to make LLMs work for you rather than substitute for you. If you would like to focus on building your own AI literacy, you might find this recent guidance from professional composition and literature organizations useful.

That said, for this class, you should refrain from using LLMs in composing your written assignments. Introductory writing courses are designed to help YOU develop as a writer and to give you some guidance for composing successfully at the university level. If you outsource your writing to an LLM, you are not developing the skills or writing abilities that this course prioritizes. The use of LLMs for drafting sentences or paragraphs will be considered a violation of academic integrity policies. Evidence of AI usage will be handled at my discretion. Depending on severity, options may include meeting with me to discuss your composition process, revising or re-writing an assignment, failing an assignment, or an Incident Report form to UW Community Standards & Student Conduct.

Initial Guidance for Instructors to Consider

In your teaching and class design, it is important to consider your perspective on the role of LLMs in writing instruction. Higher education is now responsible for building AI literacy: effective and responsible practices for using and considering Generative AI. This does not mean that we as instructors need to embrace unfettered use of LLMs in student work for every class; our approach to AI literacy may be a stance that students should, in some contexts, strengthen their unaugmented writing skills to make them more sophisticated users of digital tools. It is also important in higher education for AI literacy to be balanced with other kinds of print literacy, so it is defensible to claim that an introductory writing class should develop reading and writing skills prior to engaging with generative tools. On the other hand, it is also defensible to claim that student writing prepares students to engage with digital composition tools, and that the first-year writing classroom should teach students the critical approaches to engage with LLMs responsibly and productively. There are many valid positions for the potential role of LLMs in the writing classroom, and all   instructors will need to engage with the many  f salient pedagogical and ethical issues involved and design a policy and practice aligned with their philosophies, course goals, and students’ needs.

Issues to consider:

1) For your class, to what extent will use of LLMs be encouraged, discouraged, or forbidden and why? Is brainstorming with an LLM acceptable? Is using Grammarly for proof-reading acceptable? Is thoughtful use of prompt engineering to generate text for student revision acceptable?

2) Effective use of LLMs might include some kind of documentation. Consider whether you’d like students to indicate their use of GenAI tools with a  screenshot/readout of results, short reflection paper, a footnote, or a brief description at the end of the paper.

3) If you disallow students from generating text with LLMs, how much surveillance are you willing to provide to enforce any policies? Since English department writing programs strongly discourage use of Turnitin or other AI-detectors as problematic both for accuracy of detection and for the ways that they compromise student intellectual property, how are you prepared to engage with students about suspected use of LLMs? One possibility is to request that students submit a reflection about their writing process that might include a description of their composition process from initial learning and research to drafting, feedback, and revision. It may also be helpful to recommend that students compose in a word processing application that saves the composition history, such as GoogleDocs. Then, any conversation about their composition process could potentially draw upon evidence from composition history.

4) If you allow students to use LLMs in your class, might you build in lessons or design assignments to help them practice and navigate the ethical and effective uses of the tools in writing? How might you teach AI literacy in a way that aligns with our learning outcomes? For example, might you design a lesson focused on audience and rhetorical awareness or that helps students brainstorm or do research for assignments, create an activity that helps them understand how biases are produced in LLMs, or design an assignment that includes some intentional drafting of text by an LLMs (perhaps to consider how AI fails to understand context, audience, culturally-specific nuance, etc.)?

5) What consequences would be in place or could you imagine implementing if you believe that a student has relied on an LLM for composing in ways that exceed your permitted practices?

Sample Syllabus Language:

Depending on your comfort with ChatGPT and other LLMs, you may wish to explicitly encourage, explicitly prohibit, or allow limited, careful usage of these tools. The following language represents possible approaches to messaging with students. These statements are largely sourced from Lance Eaton’s “Syllabi Policies on Generative AI,” a document compiled leading up to the start of the 2023-24 academic year, now with over 100 crowdsourced policies.

To explore even more syllabus language, here are a few resources:

• UT-Austin Sample Syllabus Policy Statements
• Brandeis University Possible AI Syllabus Statements
• Teaching@UW guidance on AI, including some Sample Syllabus Statements

For instructors who wish to allow limited, nuanced usage of ChatGPT

Example 1, Cultural Studies course

Sarah Bunin Benor, University of Southern California

ChatGPT and other AI generators that use large language models can be useful for researching and writing papers. However, you should be aware of their limitations:

• Errors: AI generators make mistakes. Assume the output is incorrect unless you check the claims with reliable sources.
• Bias: Their output may reflect bias because the data they are trained on may reflect bias or may not include sufficient data from certain groups.
• Citation: These tools use existing sources without citation. Therefore using their outputs puts you at risk of plagiarism.

With these limitations in mind, you are welcome to use AI generators to brainstorm and refine ideas, find reliable sources, outline, check grammar, refine wording, and format bibliographies. Beyond bibliographies, you are not allowed to copy and paste material generated by AI and use it in your assignments. At the end of your bibliography, add a note indicating which AI tool you used and how you used it, including the prompt(s) you used and the date(s).

Example 2, Humanities course

Beth May, Centralia College

I am interested in YOUR thoughts, as this class is all about developing your critical listening skills, deepening your understanding of music, and exploring your own ideas.  As such, I hope that you will choose to engage with our course content and share your own words, not those of any AI.  However, I recognize that AI can be useful for brainstorming and is sometimes used for information-gathering.  If you choose to use AI in your work for this class, I ask that you do the following two things, in the following order: 1. Fact-check any information you have included from AI.  AI frequently makes mistakes, or “”hallucinates””, and even creates made-up citations.  It is in your best interest to double-check information you find via AI. 2. Disclose in your writing that you used AI, and how.

Most of the currently available AI tools are fairly recognizable in their output; there is a good chance I will recognize your use of AI, even if you don’t disclose that you use it.  There is no real reason to use AI and not disclose that information for this class – but if you use AI and do not disclose that you did so, your work will be considered an example of academic dishonesty (see below)

Example 3, History course

Chris Beneke, Bentley University

Example 4, Life Sciences course

Franklin Hays, University of Oklahoma

Use of AI tools (e.g., ChatGPT, Bard, Claude) are encouraged in this course to facilitate the student learning experience and overall productivity. However, such use should follow three clear principles: 1) any and all use should be transparent, properly cited, and otherwise declared in any final work product produced for grading or credit; 2) students are responsible for ensuring accuracy of content produced including references and citations; and 3) students acknowledge that improper attribution or authorization is a form of academic dishonesty and subject to the Academic Misconduct Code as outlined in the Student Handbook and the Faculty Handbook. All work turned into the instructor for grading is assumed to be original unless otherwise identified and cited. If there is uncertainty about any content in regard to the above guidelines, please contact the instructor to discuss these questions prior to turning anything in for grading.

Example 5, Literature course

Drexel University by Magdalena Maczynska

It’s very important that you understand exactly how AI tools work and how to use them in ways that do not limit your own agency and creativity, and do not perpetuate harmful (or false) ideas and discriminatory epistemological frameworks. To this end, we will talk about generative AI, and, more specifically, Large Language Models like ChatGPT and GPT4 in this class. You are welcome to experiment with generative AI for some assignments—I will ask you to keep track of how exactly you used AI tools, and we will debrief and reflect together to get a sense of how people are making new AI applications work for them in academic contexts. Some assignments, on the other hand, will be designated as “human only” so that you can practice essential skills like reflection or analysis, and get into a state of creative flow without algorithmic interference.

For instructors who explicitly encourage the use of ChatGPT

Example 1, History course

Christy Snider, Berry College

This class allows the use of AI tools (such as ChatGPT, Bing chat, and others) for your out-of-class assignments. AI is a new and valuable skill to master. Students who use AI correctly can become more productive, efficient, and skilled scholars. AI should not be viewed as a way to avoid learning the material, but rather as an assistant that can help you master content and produce better results.

Artificial intelligence is a rapidly evolving field that has many applications and implications for various disciplines. Different courses and instructors will have different policies regarding the use of AI tools and services for academic purposes. It is your responsibility to follow the AI policies for each of the courses that you are enrolled. Violating the AI policies of any course may result in serious consequences, such as a lower grade, a failing mark, or academic probation.

1. 1. To get high quality results from AI, you need to craft good prompts. Simple prompts lead to weak results. OpenAI the company that built ChatGPT has put out documentation on some of the best practices to use with AI chats to achieve good results.
   2. Don’t believe anything that AI tells you. If it provides a number or a fact, verify it with a trustworthy source. You will be held accountable for any mistakes the tool produces. AI fabricates or “hallucinates” seemingly credible data all the time. It can generate wholly inaccurate content that is nonetheless highly persuasive. This is especially true when asking it for references, quotations, citations, and calculations. Among the various different models, Bing tends to have the least hallucinations, primarily due to its internet connectivity. However, Bing will still lie to you; especially if you ask it to expand upon a topic for which it doesn’t have information. AI can also produced biased answers. It was likely trained on material primarily created by by straight, white, men. Your grade will be penalized for each wrong fact or inappropriate text submitted. Make sure, therefore, to closely read and verify anything AI generates to ensure that it accurately reflects the facts and your position on a topic or issue.
   3. AI is a tool that needs to be cited. If you use AI on an assignment, you need to acknowledge that in a paragraph describing how AI was used and listing the prompts used. Failure to provide this information is a violation of academic integrity. Your paragraph about AI use and prompts should be submitted a downloadable attachment in the ‘Add Comment’ feedback section of the assignment.
   4. AI is not appropriate for all situations and contexts. Be thoughtful in how you use it. AI is not a replacement for knowing and understanding the material, but it can be helpful in getting you started, learning the information more deeply, and maybe even proof-reading and improving your papers. Be aware, however, that the texts and images produced by AI prompts are currently not protected by U.S. copyright law.

Example 2, Technical Communication

Nathan Jung, University of Wisconsin–Madison (source)

Generative AI technologies will impact many aspects of the engineering workplace, including the way we craft communications. As a result, this course explores the potential for AI to refine your speaking, writing, and listening skills to better meet on-the-job challenges and strengthen your professional contributions to contemporary society. It will also explore the pitfalls of AI technology as it currently exists.

Since this course explicitly engages with generative AI platforms, there are no restrictions on the use of AI for any assignment. However, note that I do not require the use of these platforms for any graded assignment aside from the AI demos/workshops. Whatever your preference, we will explore these platforms together during class time to achieve the core learning objective of developing your capacity to effectively and responsibly use AI writing platforms.

Here are some guiding principles to keep in mind when pursuing this objective:

• Transparency builds credibility; always make sure readers know where you are getting your information from and how you are producing your content.
• Authorship entails responsibility; as an author, you are ultimately responsible for the accuracy of the information in your writing.
• AI requires human judgment; you will need to sharpen and use your own judgment on the proper times and places to use AI writing in general, the proper times and places to use AI writing in the writing process itself, and the best forms of prompting based on foundational rhetorical considerations of audience, purpose, context, and so on.
• AI raises the bar for quality communications; to elevate AI content beyond its “house style” to the level of superior writing, you will need to make significant interventions in your prompts and/or in your revisions to AI output to accommodate the nuances of argumentation, localization, and audience that distinguish good writing and speaking.

If I sense that AI platforms have negatively affecting your writing – by making it too boilerplate, for example, or by hallucinating sources – we will have a conversation about how you have used these platforms and how you might use them better in the future. And if you have questions about whether a particular use of generative writing is acceptable, bring the question to me!

For instructors who explicitly prohibit the use of ChatGPT

Example 1, Gender Studies course

Kendra Albert, Harvard University

Don’t.

Unless you receive explicit instructor permission, do not submit materials produced by large language models or “generative AI” in this course. You also may not use them for research or summarization of course materials. Doing so will be considered academic misconduct.

More specifically, we will discuss “AI” in the course, on April 1, 2024. (That is actually not a joke.) Using such tools prior to that is prohibited. If you wish to use such tools after that, I would be happy to discuss it with you. Additionally, because the course is ungraded, the use of these tools ultimately will not improve your grade and will likely jeopardize it.

Example 2, Instructional Design

Tara Perrin, Middle Tennessee State University

Use of an AI Generator such as ChatGPT, MidJourney, DALL-E, etc. is explicitly prohibited unless otherwise noted by the instructor. The information derived from these tools is based on previously published materials. Therefore, using these tools without proper citation constitutes plagiarism. Additionally, be aware that the information derived from these tools is often inaccurate or incomplete. It’s imperative that all work submitted should be your own. Any assignment that is found to have been plagiarized or to have used unauthorized AI tools may receive a zero and / or be reported for academic misconduct.

Resources to Consult

Spreadsheet of over 100 AI syllabus policies (not restricted to English Studies)

Teaching Writing Across the Curriculum with AI

Working Paper on Generative AI and Policy Development from the MLA – CCCC Joint Task Force On Writing and AI (Here’s a blog post summary of the paper)

• This paper is helpful in offering guidance for AI policy writers, offers guidelines, ethics, scenarios for classroom level, program level, and college/university level. Wish there were a lot of policy examples but the guidelines are useful for ethical and pedagogical considerations. Kinda a heuristic of questions and concerns to weigh as programs, esp do the labor you are doing here.

Exploring AI Pedagogy from the MLA-CCCC Joint Task Force on AI and Writing

• This is a helpful community culled resource with assignments for integrating AI, praxes reflections, and more. The “Reflections” tab could be helpful to explore. Could be a good resource to share with instructors.

Using Generative AI or Chatbots in the Writing Process from U of Michigan’s Sweetland Center for Writing

• A short but pretty helpful guide for teachers with links to sample policies, AI and ethics, some guidelines for thinking about writing and AI, some activities and AI tools.

AI + Teaching from UW CTL

• Some basic guidelines, sample policies, etc.

UT – Austin – this is a short run down with policy samples.

Link on the MLA and CCCC join task force on AI and writing. It has the report mentioned with guidelines but also teaching experiments, getting started on AI resources and more

From Candice, new guidance!The 4C/MLA TF on AI and Writing has now made available a student-facing handout on AI Literacy. You can find it at our TF page, at the MLA style manual page, and the MLA handbook page.

https://style.mla.org/student-guide-to-ai-literacy/

https://mlahandbookplus.org/…/Student-Guide-to-AI-Literacy

https://aiandwriting.hcommons.org/student-guide-to-ai…/