GenAI

Teaching with Generative AI Tools

Generative AI (GenAI) tools are transforming classrooms, offering innovative ways to enhance learning experiences and foster creativity. From helping us create engaging course materials to assisting students with research tasks, these tools are full of potential. In our crucial role as educators, we need to guide our students in using these tools ethically and effectively.

In Fall 2023, the College of IST charged a committee to begin exploring how AI can complement teaching while ensuring we're building our students' critical thinking and AI literacy skills. You can view the results of this committee’s exploration in the GenAI Across the Curriculum report:

Generative Artificial Intelligence Across-the-Curriculum Task Force Report

Generative AI Tools Across the Curriculum Task Force

Fall 2023 – Spring 2024

Charge

The Task Force was charged by Associate Dean for Undergraduate Studies, Dr. Sharon Huang, as follows:

Consider the problems and opportunities across our undergraduate curriculum afforded by the rise of generative AI tools. Specifically, address the following:

Given the presence of generative AI tools, how do we teach courses such that students learn the subject matter and form a strong foundation in the knowledge and skills required to be successful in their careers?
In addition to the proposed syllabus statements circulated in August, 2023 and in addition to the work that the Academic Integrity Committee is doing in Fall 2023, what recommendations does the Task Force have for tools, policies, procedures, and/or best practices that would assist instructors?
Given employers’ expectations for not only strong subject matter knowledge but also knowledge of generative AI tools and their appropriate and ethical use, how do we incorporate generative AI tools into course assignments and the curriculum in a strategic and educationally sound manner?
What are others (at Penn State or elsewhere) doing to address this question?
What recommendations does the Task Force have for addressing this question across the college and within specific programs?
- Identify which courses can benefit from Generative AI
- Criteria for evaluating courses in terms of GenAI use
- Draft modules (e.g. prompt engineering, text generation, image generation) and libraries that can be used and adapted by instructors of the courses

GenAI Literacy and Ethics

AI literacy is essential for both educators and students to navigate today’s rapidly evolving technological landscape. To build strong AI literacy, consider using a mix of resources that focus on understanding AI capabilities, ethical considerations, and practical applications.

Penn State libraries created a series of short courses useful for both faculty and students learning about GenAI:

For educators who want to learn more about GenAI, start with professional development opportunities, workshops, and AI literacy frameworks like those from UNESCO or the Penn State resources in the list below. Make sure to experiment with AI tools yourself to understand their strengths and limitations, especially for the topics and skills for the courses you teach.

When helping students to learn about effective and ethical GenAI use, incorporate structured learning experiences such as guided AI activities, discussions on ethical implications, and hands-on practice with AI-driven tools. Encourage critical thinking by analyzing AI-generated content and exploring its biases and reliability. Please see the section below for more resources on teaching ethical concerns about AI to students.

Penn State students, faculty, and staff have access to Microsoft Copilot through the University’s Microsoft contract. If you require students to use GenAI for class, they should be given an option to use Copilot. Penn State IT has developed a learning path about Copilot that you can share with students: Penn State’s Microsoft 365: Copilot Learning Path.

There are also numerous learning pathways, courses, and resources on LinkedIn Learning that faculty can share with students. Many of the learning pathways or courses provide a certificate upon completion. Students can access these resources though their LinkedInLearning account provided by Penn State. Here is an example course about prompt engineering: Introduction to Prompt Engineering for Generative AI.

There are many Gen AI tools and more coming. Please consult the AI Guidelines from Penn State’s AI Hub to know which ones are university approved and which are not.

Using GenAI in education raises ethical considerations related to bias, misinformation, privacy, and academic integrity. To foster ethical AI use in the classroom, host classroom discussions about how Generative AI is affecting your subject or industry, or even society more broadly. Consider the following when identifying topics for discussion.

Bias & Fairness: AI models may reinforce biases in the data they are trained on. Encourage students to:
- Critically analyze AI-generated content.
- Compare AI-generated responses with credible sources.
- Discuss how AI bias can affect decision-making.
Misinformation & Accuracy: AI-generated content can be plausible but incorrect.
- Teach students to fact-check AI outputs.
- Encourage students to cross-reference information with reliable sources.
Privacy & Data Security: Some AI tools store user interactions, raising privacy concerns.
- Warn students not to input personal or sensitive data.
- Use institution-approved AI tools, such as Microsoft 365 Copilot, that comply with university data protection policies.
Accountability & Transparency: Students should:
- Always disclose AI assistance when used in coursework.
- Understand that they are responsible for their submissions, even when AI is used.

Regarding privacy and data security, the University’s contract with Microsoft ensures that any use of Microsoft Copilot is secure, confidential, and not used to train the tool.

Note: ChatGPT was used to help generate these ideas which have been reviewed and revised by the IST LD team.

OpenAI. (2024). ChatGPT (Mar 24 version) [Large language model].

https://chat.openai.com/chat

For further reading about GenAI and Ethics:

GenAI and Course Design

GenAI can be a powerful tool for designing course materials when used correctly and transparently. Whether you are using GenAI to generate text for a slide deck or a Canvas Page, or generating questions or rubric language for assessments, it is important to ensure

that the generated content aligns with your learning objectives, maintains academic integrity, and is reviewed for accuracy, clarity, and bias. Here are some examples of content you can generate to review and revise for classroom use.

Lecture Notes & Summaries: Generate structured lecture notes from key concepts.
Slide Deck Content: Draft initial slide text and image suggestions.
Course Readings & Handouts: Summarize complex topics for student-friendly explanations.
Discussion Prompts: Generate open-ended questions to drive student engagement.
Case Studies & Simulations: Create learning scenarios for students to explore different perspectives.
Role-Playing Exercises: Use AI to simulate historical figures, industry professionals, or real-world scenarios.
Assessments: Create questions and rubrics.

AI-generated material should serve as a starting point, with human oversight to refine, personalize, and contextualize it for your specific audience. Transparency about AI use is also crucial as it helps students and colleagues understand when and how AI is being used in the learning process. If you choose to use AI for course design, you should be transparent with students by clearly stating that you have done it. You should cite it and include the prompt that you gave to the GenAI tool, and the citation information for the tool that you used. Here is guidance on how to Cite Generative AI Use.

By adding opportunities for students to use GenAI in your coursework you are helping students to gain a practical understanding of GenAI. One way to do this is by incorporating GenAI into new or existing assignments. The Commonwealth Campuses have developed this faculty guide to highlight some common tasks that can include GenAI skills: Addressing Learning Challenges With Generative AI: Faculty Guide. Faculty from the Schreyer Institute for Teaching Excellence have developed this resource to help guide faculty when they are considering adding GenAI to an assignment: AI, Teaching, and Learning: Questions to Consider. Here are some additional ideas for integrating GenAI into your assignments: 5 Ways to Integrate GenAI into Assignments*.

Here are some ideas on how GenAI can be used in a programming class:

Explaining Code Concepts: AI can provide step-by-step explanations of syntax, logic, and algorithms.
Generating Sample Code: Instructors can use AI to create example programs to demonstrate concepts.
Debugging Assistance: AI can help identify errors in student code and suggest possible fixes.
Develop AI Literacy: Help students critically evaluate AI-generated code for errors, efficiency, and best practices.
Compare AI-generated & Human-Written Code: Assign projects where students analyze AI-generated code and improve upon it.
Debugging Challenges: Give students AI-generated code with intentional errors and ask them to fix it.
Prompt Engineering Assignments: Teach students how to refine AI prompts to generate useful programming solutions.
Code Style & Optimization Suggestions: AI can suggest improvements for readability and efficiency.

Note: ChatGPT was used to help generate these ideas which have been reviewed and revised by the IST LD team.

OpenAI. (2024). ChatGPT (Mar 24 version) [Large language model].

https://chat.openai.com/chat

Here are some methods to use GenAI to help create assessments. As with any AI use, make sure you review and revise the outputs so that they are accurate and unbiased.

Question Generation: Use GenAI to generate quiz questions, discussion prompts, or case studies, then refine them for clarity, relevance, and alignment with learning objectives.
Rubric and Feedback Assistance: GenAI tools can draft grading rubrics and provide formative feedback suggestions, helping streamline assessment design while maintaining instructional integrity.
High Quality Example Generation: Using your project or essay prompt, ask GenAI to generate examples of projects or essays and then revise them into exemplar works to share with your students so that they know what good work looks like. This could also be an opportunity to show students the deficits of using GenAI alone—such as vague generalizations, lack of depth, or poor alignment with grading criteria. By comparing the raw AI-generated output with your revised, high-quality version, you can help students learn to identify low- and high-quality work.
Use GenAI to Review and Revise Current Assessments: Ask GenAI tools such as Copilot, or other GenAI tools to give you feedback on your assignment. Then revise your assignment (using the feedback from the tool) and use that revised version as the assignment.

Using generative AI tools as part of your course design process can make a significant impact on the quality of your class materials while saving you time. These tools can streamline the creation of course content, assessments, and interactive learning activities. Here are some resources and strategies to effectively utilize GenAI in your teaching:

Using Prompt Libraries
- Ethan & Lilach Mollick’s Instructor Prompts – These are a collection of prompts tailored for educators. They can be used to generate:
  - Scenarios, such as simulation creation
  - Role-playing exercises
  - Personalized learning activities
- Complex Prompts for Course Design – Kent Matsueda’s growing collection of prompts covering developing:
  - Course outlines
  - Creating assessments
  - Designing instructional activities
Applying Prompt Frameworks
- CREATE by David Birss: A structured method for prompt design, emphasizing Character, Request, Examples, Adjustments, Type of Output, and Extras to refine AI-generated content
- Chaining Prompts by Garima Gupta: a prompting technique to combine simpler prompts into more effective chains that help to achieve desired results iteratively
Curating your own Prompt Library
- Storage System: Choose a document type that is easy and comfortable for your style - it could be a Word or Notepad file, a spreadsheet, or a webpage.
- Structured Format: Consider using a template for each entry such as:
  - Title
  - Purpose
  - Example Prompt
  - Category or Tags (e.g. course planning, outcomes, assessments, feedback, lectures, etc.)
- Continuously Improve: Document what worked and what didn’t work. Keep notes about how you improved your prompts or what you might want to try next. List questions that you can share with your colleagues.

GenAI and Academic Integrity

Here are some recommended Syllabus policies to consider:

IST Teaching site: GenAI Tools Use Syllabus Statement
PSU Academic Integrity: Generative AI Syllabus Statements
Smeal: Sample Syllabus Statements

There are three strategies to improve transparency and communication about student GenAI use.

First, as an instructor, you must be willing to be honest and open about any GenAI output that appears in course materials. One way to do this is to cite AI use – PSU libraries has a guide for this: Citing Generative AI.

Second, if at least some GenAI is allowed in a course, students will need clear guidelines to understand when GenAI use is acceptable or unacceptable at the assignment level. One method is to embed icons that represent varying levels of acceptable GenAI use, such as these developed by Smeal: GenAI Canvas Assignment Samples or Penn State’s Commonwealth Campuses: GenAI Acceptable Use Guide. Adding transparency at the assignment level will help students know when and how to use GenAI and will help to prevent academic integrity issues.

Instructional designers from the Commonwealth Campuses and faculty from the Schreyer Institute for Teaching Excellence have developed a number of resources related to Gen AI and Academic Integrity.

Commonwealth Campuses

Schreyer Institute

Third, communicate expectations about GenAI in your class. Because of the ever-changing nature and power of GenAI, students will need frequent discussions to understand the implications of GenAI use especially as it pertains to academic integrity. Here are some aspects to consider:

Discuss acceptable and unacceptable AI usage in class (or via Canvas discussions and announcements).
Discuss how GenAI is affecting your subject or industry
Provide examples of responsible GenAI use.
Address common misconceptions (e.g., AI tools are always factually accurate. Grammarly is a GenAI tool).

If you do not allow the use of GenAI in your courses or in specific assignments but students use them anyway, that is considered an academic integrity violation by Penn State. Check the Penn State’s Academic Integrity website or the College of IST site for more information.

No detection method for text generated by GenAI is foolproof. Signs of AI-generated content can include inconsistencies in writing style, overly generic or formulaic responses, or a mismatch between a student's typical work and their submission. AI detection tools can provide some insight, but they should not be relied upon as definitive proof. If you suspect unacceptable GenAI use by a student, then you should follow the Penn State's Academic Integrity Procedures including that you review and discuss course policy and the violation with the student.

GenAI and Accessibility

: Alternative Text (Alt-Text) Generation: AI can create descriptive alt-text for images and visual content, aiding visually impaired students. If you use ChatGPT, a custom GPT created by Aimee Brown-Shadduck of IST’s Learning Design team does this well. Navigate to the "Explore GPTs" section in the ChatGPT sidebar, search for “IST Image Accessibility Assistant”, select it, and then click "Start Chat" to begin using it.

Automated Transcripts & Captions: Generate captions for lecture recordings or live discussions using AI-powered captioning tools. Faculty and students have access to Penn State’s video management system, Kaltura, which is capable of providing automated transcripts and captions for uploaded videos – check Kaltura’s Knowledge Center for how to Order Captions for a video. It is important to review any generated transcripts or captions for accuracy. The AI Hub at Penn State also notes in their Accessibility Considerations page that any AI tools used by more than one user in a class must have an accessibility review as required by policy AD69.

Teaching with Generative AI Tools

GenAI Literacy and Ethics

GenAI and Course Design

GenAI and Academic Integrity

GenAI and Accessibility

Legal Statements