Artificial Intelligence in Teaching and Learning

Guidance for Syllabus Language

Communicating when and how it is appropriate or not to use AI is critical to the success of your students. Undergraduate and graduate courses should discuss acceptable and unacceptable uses of generative artificial intelligence in the course or for particular activities/assessments. A working group of University faculty and instructional staff crafted guidelines for a syllabus statement about the use of AI. The sample statements range from forbidding the use of AI (red light) to limited use for particular activities (yellow light) and open encouragement of use (green light) with acknowledgement. Whichever path you choose, it is important to be explicit, discuss the decision with students, and post reminders in the instructions for each assignment.

Image

U of A Guidelines for Syllabus & Course Policy about AI

Can I Use AI on this Assignment? Generative AI Acceptable Use Scale

Decades of research indicate that most students cheat when they do not understand the rules, are stressed, or feel that an assessment or situation is unfair or impossible. 

Students face different policies in different courses and assignments. Clarification about expectations is necessary. This is similar to the different meanings of the phrase “open book exam.” In some one course, that means only notes are allowed. While in another course, it means students may search the internet.

Prevent situations in which a student unintentionally engages in academic dishonesty by explicitly stating and discussing what is appropriate and ethical use of AI tools in your course and/or individual assignments. Generative AI is changing rapidly and invites experimentation with new ways of transforming content and doing things, so take an approach that emphasizes clarity, transparency about sources/methods, and ethical decision-making. 

Laura Roberts at Worcester Polytechnic Institute uses Collaborate! Don’t Delegate graphics to communicate AI use expectations to students. 

It is best to approach suspected violations of your course’s AI rules in a way that supports learning. If you suspect inappropriate use of AI, your first step should be a conversation with a student. See the Dean of Students Faculty Academic Integrity Resources.

Information in prompts given to generative AI tools could be exposed. Do not use confidential information in prompts. Alert your students to privacy and security issues. When using AI tools, read warnings, privacy policies, and terms of use. Don’t copy-and-paste or upload intellectual property, including student work, without permission of the creator. 

The University of Arizona’s Chief Information Officer has decided that there are significant security and privacy concerns with Deep Seek. The U of A is blocking the use of the Deep Seek mobile application and website on University-issued devices and wired and wireless networks.

Citation and acknowledgment practices for generative AI are evolving along with the tools. If you want students to cite/acknowledge the use of AI, give instructions for how to do this in the syllabus and/or assignment(s).

Citing AI Content: When any AI-generated content is paraphrased, quoted, or incorporated into work, it should be cited; this includes AI-generated visuals or other media. If the AI tool makes its interactions shareable, tell students to include a link or URL to the original prompt. The MLA Style Center: How do I cite generative AI  or APA Style: How to cite ChatGPT have other examples. You may find yourself and your students developing citation styles specific to your course or assignments. 

Acknowledging AI Assistance: Citations don’t always encapsulate the many ways students use AI. If students use AI tools as part of their process during learning activities and assessment projects, some other ways to acknowledge AI are through reflections, research journals, cover sheets, letters to the reader, footnotes/endnotes, acknowledgment statements, or a digital survey.  Provide or create examples with your students. 

Remind students to check their work and vet AI-generated content: Generative AI tools may create inaccurate, biased, illogical, false information, and non-existent sources. Everyone should thoroughly check or investigate any AI-generated content.

UA Libraries can help: The libraries offer workshops, citation and acknowledgement style guides, and information literacy tutorials for instructors and students. Librarians also review and recommend some scholarly AI tools. We encourage you to check out UA Libraries AI LIteracy in the Age of ChatGPT

Image

AI-detection tools can be used to assist in reviewing student work, but detectors are problematic because they 

• Frequently mistake things that are human-written as AI-written 
• Disproportionately misidentify writing by multilingual learners as AI-written 
• Are not transparent about how they work 
• Do not provide sufficient evidence on their own to conclude that a student has violated course policies 

Instead of AI-detection tools, you can

• Have students share their work/thinking processes and how they used AI
• Integrate students’ use of AI, e.g., as a reviewer, editor, assistant, or tutor
• Check students’ understanding of acknowledgment/citation/disclosure practices
• Facilitate discussion of the pros and cons of AI for learning
• Implement best practices in assignment instructions, e.g., explain the purpose, why it matters, what to do, how to do it, which resources to use, with whom to work, and the evaluation criteria
• Shift to more frequent low-stakes assessment in-class / live 

If you choose to use AI detection, include a statement in the syllabus that informs students. Do this before any assessment or grading begins. If possible, include a link to the privacy policy of the tool(s) you will use. 

Use of AI-Detection Statement Example: Your work may be shared with one or more AI-detectors that are designed to predict if it was created by a generative AI, e.g. ChatGPT. 

This is a moment of tremendous opportunity for investigating and learning about AI and its impacts. AI technologies are widespread and not going to disappear. Educators are in the position of figuring out what students need to know about AI and how to use AI tools appropriately, securely, ethically, and in ways that help learning and thinking–not just in class but in preparation for work, civic and domestic life. Students are partners in figuring this out. 

Opportunities

Efficiency: These systems extend human abilities. AI assists in making and modifying lesson plans, rubrics, test questions, case studies, summaries, reports, outlines, tables, revisions, graphics, audio, and ad infinitum. Very quickly.

Creativity: Teachers, writers, and thinkers of all types describe using generative AI to brainstorm, get a better understanding of a topic, look at something from a different point of view, develop research ideas, or help them get organized. 

Learning Support: Students can interact in real time with large language model tools and get immediate feedback. As a digital learning partner, generative AI doesn’t take the place of meaningful instructor-student and student-student interaction but provides additional individualized support at any time. 

New Literacy and Job Skills: Already there is a demand for people who have the knowledge and skills to develop AI programs, critically evaluate AI technologies, communicate and collaborate effectively with AI; and use AI as a tool.

Ethical Issues and Risks

Information Security: The University of Arizona is blocking the use of the Deep Seek mobile application and website on University-issued devices and wired and wireless networks. The Chief Information Officer and IT staff find that there are significant security and privacy concerns. Assume that information shared in a prompt is visible. Personal information, e.g. social security number, home address and phone number, should not be entered. Sensitive information such as unpublished data should not be included in a prompt. 

Access Inequity: Not all tools are free; more powerful versions usually require a paid subscription. Some tools are not available in certain parts of the world. Not all generative AI tools meet disability access requirements. Students have different needs and different levels of familiarity with AI tools.

Inaccuracy, Bias, and Harm: Large language models are trained on data from the internet, digitized materials, online discussion boards, and human feedback. The racism, violence, misinformation, sexism, and ableism that exist in that data affect the output of these tools. All generated material should be fact-checked and evaluated.

Copyright & Intellectual Property: Artists, writers, and publishers are challenging AI developers in court, saying that they use copyrighted material without permission in order to train their AI programs. 

Unfair labor practices: Reports on technology companies’ business practices raise concerns about unfair labor conditions of the human workers who give feedback to train AI models.

Negative Impacts on the Environment: AI requires an enormous amount of energy and water consumption. Proliferating data centers and servers contribute to waste and heating the planet.