Skip to Main Content

Artificial Intelligence and AI Writing

This guide presents information about AI generated text and language models, how they work, how they don't. There are also resources for best practices in the classroom, and insights into AI plagiarism and detection.

Faculty Resources

As you determine your classroom / course AI policies and guidelines, consider the following questions:

  • Will you allow AI tools in your course? Which ones? Under which circumstances? Why or why not?
  • What constitutes “appropriate” and “ethical” use of these tools in your classroom? Why?
  • Is AI-generated language or code acceptable in your course? If so, how much of the text or code may be generated by AI (e.g. 20%? Less? More?)?
  • Is AI-assistance acceptable in your course? What type of assistance is ok and for which tasks?
  • What is the difference between AI-assistance and AI-generated language within the framework of your course?
  • How does AI fit in with your learning outcomes for the course? Why?
  • What are the pitfalls and limitations of AI tools in your field?
  • How would you like your students to document their work with AI tools when they submit work to you? Note that the current AI citation practices require that the prompt to the AI tool should be included in a full citation. You might also encourage students to write an “Acknowledgments” or "Works Consulted" section for their project, outlining what help they have received from whom/what.

Remember that writing something in your syllabus won’t mean that your students will read or internalize it! Please make sure that you have a conversation with your students about AI in the classroom. You may even consider having the students write or revise the classroom AI policy together.

See also this article by Matt Miller from Ditch That Textbook: "ChatGPT, Chatbots and Artificial Intelligence in Education: How to define "cheating" and "plagiarism" with AI"

Sample Classroom Policies

Please feel free to consult the Center for Faculty Enrichment (CFE) for specifics on setting your classroom policy regarding AI here at Augustana.

You may also consider consulting the Classroom Policies for AI Generative Tools document, created by Lance Eaton for the purposes of sharing and helping other instructors see the range of policies available by other educators to help in the development of their own navigation of AI-Generative Tools (such as ChatGPT, MidJourney, Dall-E, etc). 

Syllabi Suggestions

In January of 2023, the CFE issued a few general recommendations for approaching syllabi statements regarding AI. The recommendations fell into three broad categories, considered on either a class-wide or assignment-specific basis:

  • AI Use Permitted
  • AI Use with Restrictions
  • AI Use Prohibited
Approach / Philosophy Example Statement(s)***
AI Use Permitted

Statement 1:

"Learning to use AI is an emerging skill that we will explore together in this class. I expect you to use AI in this class. In fact, some assignments may require it. [suggestion: include a list of specific tools if relevant] 

However, you should be aware of the limits of AI:

  • AI is a tool, but one that you need to acknowledge using. Any ideas, language, or code that is produced by AI must be cited, just like any other resource. [sample suggestion: Please include a paragraph at the end of any assignment that uses AI explaining what you used the AI for and what prompts you used to get the results.] Failure to cite sources is in violation of the Honor Code at Augustana College.
  • Don’t trust anything AI says. If it gives you a number or fact, assume it is wrong unless you either know the answer or can check in with another source. AI works best for topics you understand.
  • If you provide minimum effort prompts, you will get low quality results. You will need to refine your prompts in order to get good outcomes. This will take work.
  • Be thoughtful about when this tool is useful. Don’t use it if it isn’t appropriate for the case or circumstance.

If you have any questions about your use of AI tools, please contact me to discuss them!" (Adapted from Macalester College)

Statement 2:

"Use of AI tools, including ChatGPT, is permitted in this course for students who wish to use them. To adhere to our scholarly values, students must cite any AI-generated material that informed their work (this includes in-text citations and/or use of quotations, and in your reference list). Using an AI tool to generate content without proper attribution qualifies as academic dishonesty." (From UMass Amherst Center for Teaching and Learning)

AI Use with Restrictions

Statement 1:

"Generally speaking, you are not authorized to use artificial intelligence engines, software, or artwork generating programs (or similar) to produce work for this class EXCEPT on assignments that I have identified and for which you will have received significant guidance on appropriate use of such technologies. I will provide more information about the specific assignment when the time is appropriate in the course. You may not, however, construe this limited use as permission to use these technologies in any other facet of this course." (Adapted from Colorado State University)

Statement 2:

"In this course, you may use AI tools (such as ChatGPT) to help you generate ideas and to brainstorm. However, you should note that the material generated by these tools may be inaccurate, incomplete, or otherwise problematic. Beware that overuse of AI may stifle your own independent thinking and creativity, and use any tools (for generating text, code, video, audio, images, or translation) wisely and carefully.

You may not submit any work generated by an AI program as your own. If you include material—including both ideas and language—generated by an AI program, it should be cited like any other reference material, both in this course and at Augustana College in general. If you have any questions, please feel free to contact me." (Adapted from Macalester College)

AI Use Prohibited

Statement 1:

"Using AI can impede your learning. The assignments in this class challenge you to develop creativity, critical-thinking, and problem-solving skills that AI does not have. Using AI technology could limit your capacity to do this type of work, and as the instructor, I urge you not to miss out on the educational opportunities that this course will provide. Work submitted by you for this class should reflect both your own ideas and your own language and you should properly cite any resources you have consulted. If you have any questions about citation or about what constitutes academic honesty in this course or at Augustana College in general, please feel free to raise these questions in class and/or contact me to discuss your concerns." (Adapted from Macalester College)

Statement 2:

"A Note on AI: This class is specifically a space for learning and practicing invaluable writing and researching processes that cannot be replicated by generative artificial intelligence (AI). While the ever-changing (and exciting!) new developments with AI will find their place in our workforces and personal lives, in the realm of education, this kind of technology can counteract learning. This is because the use of AI diminishes opportunities to learn from our experiences and from each other, to play with our creative freedoms, to problem-solve, and to contribute our ideas in authentic ways. In a nutshell, college is a place for learning, and this AI simply cannot do that learning for us. Academic integrity plays a vital role in the learning that takes place in this class, and submitting work as your own that was generated by AI is plagiarism. For all of these reasons, any work written, developed, created, or inspired by generative artificial intelligence does not lend itself to our learning goals and will be considered a breach of ethical engagement and the Honor Code at Augustana College." (Adapted from Colorado State University)

***Note that these samples have not been specifically endorsed by CFE or the Honor Council at Augustana College. Please use them as examples from which to start drafting a policy specific to your course. Again, feel free to consult with the CFE as you do so.

Tools to detect machine generated text are relatively new. These tools are being developed, released, and improved all the time, but should not be considered reliable, with many of the tools flagging both false negatives and false positives. Longer text samples may usually work better, but some tools also limit the amount of text they test. Because of the limitations of these tools, there is simply no foolproof test to determine with complete certainty if a text was generated by an AI chatbot.

A few examples of attempts at detection tools include:

Instead of relying on detection after the fact, you might consider alternate ways of structuring assignments that scaffold the drafting process into what gets turned in as a means of "showing the work." Obviously that may not be viable for every assignment, but consider at least the following couple of examples:

  • For written assignments, you might require the use of Google Docs or a Microsoft Word document that includes all of the track changes data. If an essay appears fully written in its final form with no edits along the way, that's honestly probably a better indicator that something was written by AI than any of the detectors out there. Note that the "track changes" feature is enabled in both Google Docs and Word by default, so don't let a student try to tell you that they forgot to turn the feature on.
  • For graphic design assignments, you might require students to submit all of their layers rather than a single flattened final image.

Other red flags that you might consider:

  • Repetitive answers: ChatGPT can answer the same question in different ways, but often not very different; similar answers from several different students may mean they are using AI generated content.
  • Fluffy verbosity: It has a tendency to use filler language.
  • Overly biased toward neutrality: It provides factual (or at least factual sounding) answers but refrains from offering opinions or judgments.
  • That part about "factual sounding": It can write plausible sounding answers that are in fact counter to fact.
  • Fake citations: For an example of false facts, it can fabricate citations, such as using a real author's name and a real journal title but make up a title of an article the author plausibly could have written but never actually wrote.

Designing "AI Proof" Assignments

Consider structuring assignments so that students turn in versions of their work that show their work. The following articles build on a few tips mentioned in the previous "AI Detection" tab of this page:

The following curated lists of classroom ideas related to AI draw from the activities in the Creative Ideas for using AI in Education crowd-sourced resource.

General Activities

Visualize a Metaphor Take a visual metaphor from a class reading and generate an image of it using Dall-E. Have students discuss the image and how it shifts their understanding of the text or idea. Alternatively, have students write a visual metaphor that reflects their response to any type of text or idea. Generate an image of that metaphor and discuss to see if it leads to new insights or responses.
Scenario-Based Assessment Have ChatGPT create a series of scenarios related to your course content. Maybe the Bot plays the role of a client to a business who needs a marketing strategy and you ask it to come up with the specifics of what needs marketing. Or perhaps the Bot plays the role of a politician asking a policy expert to come up with recommendations for an upcoming vote. Think creatively and use the scenarios for assessments in your course. Try using the phrase “act as an expert” in your prompt to AI to get better results.
AI Gameshow Use ChatGPT or other AI-tools to write questions for a gameshow related to your current class topic. Think about what format you would like the questions to be (Jeopardy? Who wants to be a millionaire? Family Feud?) You can use the resulting questions for a review at the end of the unit or as a quick warm-up.

AI Conversations

Use AI as a conversation buddy to sharpen question-posing and critical thinking skills. Maybe each student starts with the same general prompt and then poses follow-up questions to the tool to see what happens. What type of conversation develops?

 “To create opportunities for conversational learning using AI, in this case ChatGPT or similar tools, as a conversation buddy. Could this type of conversations help us develop and sharpen our socratic questioning techniques, and open up to diverse perspectives? Could it be a way to practice active listening, critical reading and deep reflection? Could it open our minds to explore new connections and possibilities through questions? Could our own questions help us question our own beliefs, positionality and challenge our own assumptions and study contradictions? Could this approach help us develop our skills in creating safe and non-judgemental spaces for conversations in the spirit of Socratic questioning (Paul & Elder, 2007)? Are responses only as good as the questions we ask? And what are the implications if this is the case?”

Of course, this could also be done in a language classroom in another language.

(From Creative Ideas for using AI in Education, Slide 20)

Talk with an Historical or Famous Character

Similarly, you could have students hold a conversation with a simulated historical figure or a famous personality.

“Students will go to either Historical Figures Chat or Hello History or Character AI and chat with a historical figure. Then they will reflect on the chat. Did the AI make any factual errors? Did the AI use the same kinds of language? Did the AI hold the same opinions? If not, where can you find sources that prove your point? This will send students to all sorts of primary and secondary documents as they reflect on their experience.”

(From Creative Ideas for Using AI in Education, Slide 39)

Writing Activities

What is writing?

Have students discuss what writing is. What is their personal definition of writing? Is it simply putting words on a page? Does the definition vary from context to context, from genre to genre? What is the connection between writing and thinking?

The goal is to have a discussion about the fact that writing is (critical) thinking, that the two are inseparable. Our thoughts only become our thoughts through the process of articulating them, and for that reason, ChatGPT and other AI tools are not capable of this type of writing. This could open up a broader discussion about the goals for your course/assignment and the role that AI may or may not play in it.

Writing Samples

Put your prompt through ChatGPT and see what it produces. Tweak and play with it. Then give the sample to your students and have them critique it. What does it do well? What is missing? What doesn’t it do well? Use your grading criteria to evaluate the paper: what grade would it receive? Why?

Variations on this: you could tell the students in advance that the paper is AI-generated OR you could not tell them until after they’ve analyzed the sample. You could have students compare the AI-generated sample with a human-generated sample and identify the difference (again, with or without telling them which is which).

AI Proofreading

Have your students ask ChatGPT to proofread a sample essay. What feedback does it provide? What is useful and what is not? How do the students feel about the suggestions that ChatGPT makes? Poll the class: is ChatGPT a useful resource for proofreading? Why or why not? How might the class response affect the class policy on using AI for proofreading?

Note the data-privacy concerns inherent in this activity. We certainly do not want to require students to add their personal writing/data to ChatGPT, but this exercise could provide a useful starting point for discussion.

Stylin’ and Profilin’

“Use AI-generated text to delve into the specifics of literary styles by analyzing its approximation of different authors’ writing. For example, asking the AI to generate some text in the style of Virginia Woolf, Cormac McCarthy, and David Foster Wallace. AIs such as ChatGPT aren’t “intelligent” in the way some people think. But it is (already) good at creating pastiche texts in the style of particular authors. Considering “why” the AI emulated the style (sentence formation, word selection, punctuation and grammar, etc.) as it did requires learners to undertake deeper consideration of the works of the authors in question to pinpoint style particulars, and where the parallels are found in the authors’ work. It is at least as important to consider where ChatGPT gets things wrong.”

“AI-generated builds on traditional compare/contrast/analysis by providing another layer for considering the construction and effects of writing style. There are many possibilities for building on this kind of activity, including using different genres, having students write in different styles, or even asking them to emulate what they think the AI might produce before asking it to do so.”

(From Creative Uses for Using AI in Education, Slide 36)

The following graphic comes from Matt Miller at (not all ideas are necessarily endorsed, but hey, we're brainstorming here):