Why You Can’t Quality-Control What You Don’t Understand

By Jennifer Zaylea, Assistant Director of Emerging and Digital Technologies

Throughout the teaching profession, we keep hearing that “the AI can do it.”  Some instructors are even experiencing a sense of existential dread at the prospect that they no longer need to teach our students certain skills—in some cases, skills previously thought to be crucial to their field. 

But is that really true? Is generative AI (GAI) perfect? Does its knowledge exceed the human knowledge that created it? These questions are stumbling blocks for faculty walking into class every day. Students often don’t see how GAI might be undermining their learning, and some faculty are beginning to wonder whether foundational skills still matter at all. 

Consider a simple example. A GAI tool can write code for my project, but that doesn’t mean the code will produce the output I want across multiple runs. I still have to verify it against the demands of the project. If I can’t understand a line of code, I can’t prompt the tool in a useful way to identify and resolve an error, and I can’t responsibly move forward. Without foundational knowledge, we cannot examine GAI output in any deep or meaningful way. Instead, we become subjects of automation bias, the tendency to trust automated output over our own judgment. 

An Old Problem with a New Face 

The idea that automation diminishes human capability is not a new complaint. It is a well-established principle of human-system interaction. In her seminal 1983 paper, “Ironies of Automation,” cognitive psychologist Lisanne Bainbridge described the paradox: the more advanced an automated system becomes, the more crucial the human operator’s skills become. When the system encounters an anomaly or fails, a human must step in. But if that person has not been practicing manually then their skills have eroded and they no longer have the expertise needed to fix the problem. 

Engineer and former Chief Scientist of the U.S. Air Force Mica R. Endsley extends this argument to AI in her article “Ironies of Artificial Intelligence.” In higher education, we see the same pattern of automation bias but frame it as deskilling. Skimming or lightly editing an AI-generated paragraph may feel efficient, but it bypasses the cognitive effort that drives deep, transferable understanding (Bauer et al., 2025). Over time, this weakens the development of disciplinary knowledge and makes it harder for students to recognize when an AI output is biased, superficial, or simply wrong. 

Teaching Foundations in the Age of GAI 

None of this means students should be kept away from GAI. It means we, as faculty, still need to instill the foundational knowledge of our disciplines. The question of the hour is how to do that without encouraging over-reliance on the tools. 

One option is to limit some assignments to in-person, handwritten work. Used exclusively, though, this approach eventually becomes a disservice to students, who also need to learn how to evaluate both the input and the output of GAI. A more balanced strategy is to teach GAI fluency directly, using the tool as a thinking partner or tutor. This can reinforce the foundational knowledge you present in class by prompting students to connect concepts to their lived experiences. 

You might also have students work backwards from a finished product, such as a written draft of a literature response or a completed coding project. Ask them to dig into what makes up the final piece by analyzing its underlying elements, like citations and code syntax. Teach them to verify output as they work through the process. Finally, help them identify moments when they accepted an incorrect GAI output and understand how they missed it. (Hint: they probably lacked the foundational knowledge to catch the error.) 

Talking with Students About Why It Matters 

Understanding why foundational knowledge matters is only half the challenge. Faculty also need language that resonates with students in the moment. When students question the value of a skill GAI can perform, the goal is not to dismiss the concern. It is to help them see the difference between completing a task and developing the capacity to do meaningful work. 

Students may argue that it takes too long to learn the foundations, or protest, “But I’ll never use this skill again. Why should I have to learn it?” That reaction is understandable, but the need remains, and productive struggle is where real learning happens. The goal is often not mastery of one specific task but the development of ways of thinking that transfer to new situations. Problem solving, analytical reasoning, evidence evaluation, and communication stay valuable throughout life, and broad, well-developed knowledge structures often pay off in unexpected ways. 

According to How People Learn II, expertise develops when learners build interconnected knowledge structures that help them recognize patterns, organize information, and reason through unfamiliar challenges. Experts are distinguished not simply by what they know, but by how they connect and apply it. So try reframing the conversation. Move away from “Will I perform this exact task in the so-called ‘real world’?” and toward a question the research supports: “What capabilities is this helping me develop that will transfer to new situations?” 

References 

Bainbridge, L. (1983). Ironies of automation. Automatica, 19(6), 775–779. https://doi.org/10.1016/0005-1098(83)90046-8 

Bauer, E., Greiff, S., Graesser, A. C., Scheiter, K., & Sailer, M. (2025). Looking beyond the hype: Understanding the effects of AI on learning. Educational Psychology Review, 37(2), 45. https://doi.org/10.1007/s10648-025-10020-8 

Endsley, M. R. (2023). Ironies of artificial intelligence. Ergonomics, 66(11), 1656–1668. https://doi.org/10.1080/00140139.2023.2243404 

National Academies of Sciences, Engineering, and Medicine. (2018). Knowledge and reasoning. In How people learn II: Learners, contexts, and cultures (Chapter 5). The National Academies Press. https://doi.org/10.17226/24783 

 

Jennifer Zaylea, MFA, serves as Assistant Director of Emerging and Digital Technologies at Temple University’s Center for the Advancement of Teaching.

Five Temple Faculty Share Their Work with A.I. 

By Jeff Rients

If you have been wondering what a thoughtful use of A.I. might look like in your course, five Temple colleagues have some ideas to share with you. Their projects grew out of the Spring 2026 Faculty Learning Community on Teaching and Learning with A.I., and you can now explore their work on the CAT website. Think of the collection as a colleague opening a course folder and talking you through what’s inside. You get an activity and, in several cases, an account of the thinking behind it. 

Jennifer Zaylea, the CAT’s Assistant Director of Digital and Emerging Technologies, and I had the privilege of facilitating the group, which brought together faculty from business, English, and public health. A faculty learning community gives colleagues time to investigate shared questions and develop something useful for their teaching. Our questions included how A.I. might support course goals, what students need to learn about its use, and where it might interfere with the learning we want to happen. The resulting materials give us something concrete to examine. 

Gary Blau, in the Fox School of Business, used A.I. to develop role play activities for his Critical Skills for Effective Managers course. Students practice delegating work, giving performance feedback, handling conflict, and asking for a raise. In that last scenario, the employee has accomplishments to point to and the manager has a limited budget. Students must work through the conversation (preferably before an actual paycheck is involved). Observers add another perspective to the discussion afterward. 

Rob Faunce, in English, combines mindmapping and A.I. to help students prepare for a First Year Writing paper. His activity brings together research, essay planning, and work with Claude. Robb’s activity employs reflection to help students examine the tool’s possibilities and limitations. His account keeps returning to students’ critical thinking and the importance of their own voice. What can the machine contribute, and what does the writer need to bring? 

Ryan McKee, in public health, designed a lesson introducing A.I. as a thinking partner for community health workers. Students first plan a client visit themselves, then compare their ideas with A.I. suggestions. They identify useful contributions, cross out unrealistic or irrelevant advice, and add ideas the tool missed. The exercise also asks them to consider privacy and the limits of A.I.’s knowledge of a community. Students learn that a plausible suggestion still needs someone to judge whether it makes sense here. 

Ksenia Power, also in public health, used A.I. to help with lesson planning. Her macronutrients lesson includes a “Diet Detective Challenge” in which students examine sample diets and propose changes. Her accompanying reflection describes the work of evaluating and revising generated material to fit her course goals and students. For faculty curious about A.I. in course preparation, this offers an example of how an instructor develops an initial draft into something suitable for a particular class. 

Christopher Wheldon, our third public health professor, asks students to teach a chatbot to apply a theoretical framework. They first complete an analysis themselves, explain their reasoning, and turn those decisions into instructions for the chatbot. Then they test, compare, and revise. His materials include a public health example and a template that colleagues can adapt to other disciplines. How clearly do students understand a concept? Asking them to explain its application, including the ambiguous cases, forces them to examine and refine their own learning. 

The faculty reflections deserve your attention, too. McKee, for example, describes finding A.I. useful for generating activity ideas while remaining cautious about its accuracy and protective of his responsibility to understand what he teaches. As intellectually serious academics, we can investigate a possibility while keeping our reservations. Indeed, those reservations may help us ask better questions about an activity’s design. 

Across these projects, a common concern emerges regarding the intellectual work students and faculty still need to do. What must students understand to evaluate an answer? What should they attempt before opening the chatbot? How will we know whether the activity helped them learn? These are questions we can bring to our own classroom AI experiments. Of course, we should be asking other questions as well, not the least of which is, should I be using A.I. with my students at all? 

We thank Gary, Rob, Ryan, Ksenia, and Christopher for sharing their work. Visit the collection, choose an example, and consider what you would keep or change for your students. For help thinking through an adaptation, contact the faculty member or reach out to Jennifer Zaylea or me at the CAT. Our Faculty Guide to A.I. offers further resources. 

Jeff Rients is Associate Director of Teaching and Learning Innovation at Temple University’s Center for the Advancement of Teaching. 

Disclosure: ChatGPT assisted in drafting this post. 🙂

The New Grammarly: Not the Same as the Old Grammarly

By Jeff Rients

As our response to the academic integrity challenges produced by large language model A.I.s has developed and we have more and more conversations with faculty, we’ve noticed a pattern emerging. When instructors sit down with students to discuss unauthorized use of A.I. (as we described in this post), many productive conversations occur, allowing for both us and our students to learn and grow. However, some of our students, when suspected of unauthorized A.I. use, will plead they “only used Grammarly to check grammar.”

If you are not familiar, Grammarly first appeared in 2009 as an online tool to assist students with grammar and spelling. Given its longevity, you may have some students who have used Grammarly for virtually the entirety of their academic career. Running written schoolwork through Grammarly prior to submission may feel to them like a perfectly normal part of their workflow as students. Additionally, many international students regularly use Grammarly to help ensure that their written English is understandable. So when the “Grammarly excuse” is offered, some students are being truthful with us. They aren’t necessarily trying to cheat the system.

However, in 2023 Grammarly began incorporating a series of A.I.-powered tools, allowing it to generate text from a prompt, similar to other A.I. tools such as ChatGPT. Thanks to these new features, even if a student only used Grammarly to brush up their grammar, use of this tool creates the appearance of impropriety. (Of course, if you are teaching a writing class or other course where grammar is being assessed, then ANY use of Grammarly is problematic.)

Furthermore, Grammarly has not attempted to hide the fact that it now employs A.I. features. A Google search for Grammarly shows how the site wears its A.I. on its sleeve:

And the front page of the site is no less subtle:

This overt presentation makes it easy for us to file Grammarly alongside other A.I. tools. But remember: students were making use of Grammarly long before ChatGPT first arrived on the scene. Even as some students say things like they “only used Grammarly to check grammar” (A clear admission of knowledge that Grammarly has AI functionality, otherwise why mention it?) they may genuinely believe that it somehow still doesn’t count as unauthorized use of A.I.

Obviously, this is not an entirely rational proposition, but try to keep in mind that this tool may have gotten them through many writing tasks in their K-12 education. If we are to overcome students’ predisposition to carve out a special exception for Grammarly, we must be very explicit. As we review our A.I. policies with them, we should mention Grammarly by name, including Grammarly in the statements on our syllabi, etc. Students need to know (and they need to be reminded) that you consider Grammarly an A.I. tool (because it is) and that the “Grammarly excuse” won’t work.

However, this doesn’t necessarily mean that Grammarly can’t have a place in a student’s toolbox. It’s a very useful tool for refining written communication, one that could serve them well in their careers. The issue is the same as with any kind of tool that allows us to make mental shortcuts: do students know enough to be able to judge when the shortcut is working and when it is leading them astray? You may want to try it yourself, just so that you are better aware of its capabilities.

To this end, consider allowing Grammarly to be used in your classroom, but with a structure that allows both you and your students to see what is happening.  Consider the following plan:

      1. Students write first drafts of papers outside of Grammarly and submit them for simple complete/incomplete grading and clear feedback from you on important content and structural issues. You give no specific feedback on grammar other than to point out parts of the draft that you think need the most polishing. Detection of A.I. usage at this point would be grounds for sanctions.
      2. Students revise based upon your feedback. For the revised draft they are allowed to use Grammarly to clarify their writing.
      3. Students submit revised drafts, as well as memos outlining a) how they incorporated your feedback from step 1, b) the three most important changes suggested by Grammarly, and c) what they learned by revising their papers with the tool. The final grade for each paper is based upon both the paper itself and the learning documented in the memo.

Asking for the three most important changes forces students to review the changes and make a judgement call about which ones were critical to the writing task, while explicitly asking what they learned from the process demands they consciously interrogate whether the tool is aiding or hindering their education.

As with nearly all important considerations in the classroom, there are no one-size-fits-all solutions to the impact of A.I. on learning. But making sure you and your students are on the same page on how you want them to use A.I. in your course is the best first step you can take.

Akuchie, Michael. “GrammarlyGo: Everything You Need to Know About the AI Writing Assistant.” ScreenRant, 16 Mar. 2023. Accessed 12 November 2025.

5 Ways AI Can Help You Meet Title II Digital Accessibility Requirements

Dana Dawson, Ph.D

5 ways ai can help you meet title II Digital Accessibility Requirements

If you’re feeling overwhelmed by accessibility guidelines, you’re not alone! While we strongly recommend that you take a CAT training on how to meet these requirements, in this post, we outline five ways generative AI can be used to support your efforts to meet Web Content Accessibility Guidelines (WCAG) 2.1. Whether you’re already using AI tools or just getting started, these five strategies will help you more efficiently create accessible course materials or remediate inaccessible materials.

1. Identify WCAG 2.1 Compliance Issues

The Web Content Accessibility Guidelines (WCAG) 2.1 provide the technical standards for digital accessibility, covering everything from color contrast to video transcripts. Understanding how these guidelines apply to your specific course materials can be complex.

How to implement: Describe your course setup, the item in question, the application you’re using, and any other pertinent information, and ask a generative AI tool to explain which guidelines are most relevant to your content. The AI can break down technical jargon into plain language and prioritize which issues to address first.

For example, “I use Notability to teach anatomy. Thinking of the WCAG 2.1, what barriers might this create for students? Briefly, what would be needed to remediate inaccessible course materials? Can you summarize the issues with the visual (again, according to WCAG 2.1)?”

2. Generate Alt Text for Images

Every image in your course materials that conveys information needs alternative text (alt text) so that students using screen readers can also access the information. Writing descriptive alt text for dozens of images can feel daunting, but AI can help.

How to implement: Tools like ChatGPT, Claude, or Microsoft’s built-in AI, Copilot, can analyze images and suggest alt text. Upload or paste in an image and use a prompt such as, “Create alt text for this image.” The AI will provide a description you can review for accuracy and refine based on the image’s context in your course. Context matters for alt text; before copying and pasting the alt text into your Canvas page, PowerPoint slide, etc., consider your purpose in using the image and whether the generated text conveys the information that would be communicated to an individual visually perceiving the image. If it does not, reprompt the AI, adding contextual details (“I am using this image for X purpose / to teach X concept. Create meaningful alt text for the image.”) or edit the generated alt text to suit your purpose. Remember that if the image does not convey any meaningful information, you can “mark as decorative.”

For PowerPoint users, Microsoft’s accessibility checker now includes AI-powered alt text suggestions but these may be inaccurate and must always be reviewed.

3. Transform Complex Content into Plain Language

Clear, straightforward language benefits all students, but it’s especially crucial for students with cognitive disabilities, learning disabilities, and English language learners. Courses frequently include complex terminology and concepts that we expect students to know as part of the learning goals for the course and that we teach and reinforce over a semester or course of study. However, there are also many cases where we use unnecessarily complicated descriptions or instructions that can and should be simplified.

How to implement: AI excels at simplifying complex text while preserving meaning. Take a dense paragraph from your course materials or instructions for an assignment and ask an AI to “rewrite this in plain language for a X-year student” or “simplify this explanation while keeping the key concepts.” While simplification is valuable, don’t sacrifice necessary academic vocabulary. Ask the AI to maintain discipline-specific terms but define them clearly on first use.

You can also use AI to check readability. Paste your text and ask for a readability analysis with suggestions for improvement. The AI can identify jargon, overly complex sentences, and passive voice that might confuse readers.

4. Analyze and Improve Content Structure

Proper document structure—using heading hierarchies, short paragraphs, and clear organization—makes content accessible to screen readers and easier for everyone to navigate, especially those using magnifiers..

How to implement: Upload a document to an AI tool and ask it to “analyze this document’s structure and suggest improvements for accessibility.” The AI can identify heading hierarchy problems (like jumping from H1 to H3), flag “walls of text” that need breaking up, and suggest where to add subheadings for better navigation.

For online course pages, AI can review your content and recommend structural improvements: “This page has 2,000 words in a single block. Consider breaking it into sections with clear headings: Introduction, Learning Objectives, Required Readings, and Assignment Details.”

5. Develop an Action Plan for Accessibility Improvements

Once you’ve identified accessibility barriers, creating a realistic remediation plan is crucial. AI can help you understand not just what needs to change, but why it matters for your students.

How to implement: Use an AI assistant to create a prioritized action plan. For example, you might say: “I discovered that my syllabus PDF isn’t tagged properly, my lecture slides lack alt text, and my videos don’t have captions. Help me create a plan to address these issues over the next month.”

The AI can suggest a logical sequence, estimate time commitments, and explain the impact of each improvement. More importantly, you can ask follow-up questions like “Why does PDF tagging matter for screen reader users?” This deeper understanding helps you make more informed decisions about future content creation.

Getting Started Today

You don’t need to implement all five strategies at once. Consider course content that is most likely to present digital accessibility barriers to students and start there. For additional guidance on meeting WCAG 2.1, visit our Accessibility Resources page.

Remember that AI tools are assistants, not replacements for human judgment. Always review AI-generated content to ensure it meets your educational goals and accurately represents your material. As you become more comfortable with these tools and with aspects of web content accessibility, you’ll develop workflows that make accessibility feel less like an obligation and more like a regular part of your course design process.

Your students—all of them—will benefit from your efforts to create more accessible learning environments. 


Written with input by Claude, Anthropic, 30 Oct. 2025, claude.ai.

How to Talk to Your Students About A.I.

How to talk to your students about AI with CAT logo and Edvice Exchange

One thing many of us struggle with is how to talk to our students about A.I. These tools are useful and fast becoming ubiquitous, but can short-circuit student learning and undermine the development of critical thinking skills. Faculty member Graham Dobereiner made a video for his general chemistry students explaining how they might benefit from A.I. tools and what pitfalls to avoid.

There are multiple approaches to AI in the classroom and Dr. Dobereiner’s take won’t fit every learning environment. But we wholeheartedly agree with his laser-like focus on how these tools impact student learning. Whatever your A.I. policy, that should be your primary concern.

Temple faculty who want help shaping classroom AI policy or developing video materials for students, please email us at the CAT. We’re here to help!

Navigating AI: Essential EDvice Exchange Reads for Fall

by Dana Dawson, Ph. D.

With the beginning of the fall semester steadily approaching, you may be pondering how you will address the use of generative AI in your courses. To help with your decision making, AI student guideline drafting as well as activity and assessment designing, we encourage you to take a look back at EDvice Exchange posts on the topic of planning for AI use in your classes. 

A Survival Guide to AI and Teaching

Our series “A Survival Guide To AI and Teaching” featured posts on everything from what generative AI is to the ways these tools will impact equity in education. 

Faculty Adventures in the AI Learning Frontier

Our spring 2024 series “Faculty Adventures in the AI Learning Frontier” showcased how Temple faculty members used and talked about AI in their classrooms during the previous semester.

In the final post in the series, Michael Schirmer, who teaches in the Fox School of Business, shared his experiences with generative AI both with his students and as part of his personal scholarly practice.

Using PI to Manage AI

Our series “Using PI to Manage AI” considered the most fundamental element of addressing AI in our courses: sound pedagogy. Posts in this series focused on evidence-based ways of designing assessments of student learning that encourage academic honesty, motivation, and a desire to learn.

CAT Tips Season 5 – Generative AI Tools

Finally, our most recent series of CAT Tips, short videos offering teaching tips and suggestions, focused on how generative A.I. tools can be used to support student learning.

CAT staff are available for consultation throughout the summer. To schedule an appointment, visit our consultation booking page. The CAT’s Ed Tech labs are also open Monday through Friday, 8:30-5 if you’d like to drop in and chat with a staff person about AI tools (or any other educational technology question). 

Our Faculty Guide to A.I. webpage features helpful information and resources, including information on Temple’s policy. 

We also welcome you to sign up for one of our pre-semester AI workshops.

AI or Nay? Deciding the Role of Generative AI in Your Classroom

Thursday, August 8, 2024, 11:00AM-12:00PM via Zoom

Register for this workshop

Generative AI tools to help with research, writing, ideation and creative work are now part of our educational landscape and cannot be ignored, but you may be feeling unsure how to address them in your classes. Should you encourage their use but place parameters on how they are used? Will allowing students to use AI tools reduce the efficacy of your classes? Or is AI use now an essential skill for our students? In this workshop, we will explore the implications of allowing students to use generative AI tools and help you work through the question of whether and how you will permit their use in your classes.

AI Assignment Re-Do Bootcamp

Register for this workshop

Monday, August 12, 2024 and Wednesday, August 14, 2024, 1:00PM-4:00PM, Tech 109

We have been learning more about generative AI tools such as ChatGPT and what it means to teach in a world where generative AI is available to us and our students. In this 2-day intensive workshop, faculty will collaborate to revise and/or develop assignments for the AI era. You will learn how to intentionally leverage AI tools for learning and development as well as how to modify assignments to make them more AI resistant. You will also learn a framework for discussing the assignment and the role of AI in your classroom with students. This is an opportunity for you to develop the best assignments you’ve ever used in your classroom. Bring an assignment you already use and stride boldly into the future with us!

AI Sandbox

Register for this workshop

Wednesday, August 14, 2024, 12:00PM-1:00PM, Tech 109

Have you been putting off familiarizing yourself with AI tools? Or have you tried some of the more commonly available apps but would like to learn more about the wide variety of tools that are available? Join us for a hands-on exploration of the AI tools that are changing the way we live, work and study.

An End-of-Year Message From Me (and ChatGPT)

Stephanie Fiore, Ph.D.

As I faced the blank page in writing this blog post, I struggled to decide what I could say to you, my dear colleagues, that would express both a sense of gratitude for the work you have done this (not so easy) year, a reminder to take some restful time for yourself that will recharge your batteries, but at the same time, urge you to prepare more fully for teaching in the era of generative artificial intelligence. So I went to ChatGPT and asked it to help me. I chuckled as I did it because I love to write and I still find it amusing to play with GenAI tools, but also, I kid you not, it was helpful in getting me started. So I will pull out some of the words it suggested to craft for you a heartfelt message of gratitude, care, and support.

Celebrate our shared achievements and the resilience that defines our community. Each of you has played a crucial role in fostering an environment of inquiry and excellence. ChatGPT isn’t wrong that you have yet again shown great resilience as we cope with shrinking budgets, a changing student body that poses new challenges, and the advent of generative AI tools in the teaching and learning landscape. And yet, the excellence of what you provide for our students is still intact, a testament to the care you demonstrate in the teaching work you do. Temple has recently admitted the most diverse student body we have ever had, which required faculty to discover new ways of teaching that can better support all students to meet their potential. I have been impressed with how many of you have responded by engaging with us to rethink your teaching and retool your practices. And, of course, there’s generative AI. So many of you sought our answers to whether and how you might harness the power of AI or simply think about teaching when AI is readily available. Students at Temple are fortunate to have caring and resilient faculty who continue to navigate these new challenges in order to meet our mission of educational excellence.

Rest and recharge. Wellness is integral to our professional effectiveness and personal happiness. You’re so right, ChatGPT! The time you take to rest and engage in restorative activities is time wisely spent. Having spent the majority of my career as faculty, I know all too well the never-ending work. My hairdresser once jokingly asked where my grading work was when I showed up once without a red pen and papers to grade in hand. I took my work with me to my kids’ swim practices, on longer car drives, and anywhere I thought I could squeeze in some stuff. And summer is sometimes the only time available to get research or creative work done when your teaching load is heavy. We are so wired for productive energy that we sometimes forget that rest is productive too. Enjoyment is restorative and gives us the oomph to keep working later on. So please do take some time to rest, recharge, and find your happiness. Resist the urge to fill every moment with chores.

Let’s dedicate some time to exploring AI tools and platforms. I agree wholeheartedly, ChatGPT! When you are again restored and ready to explore directions for the fall semester, I believe it is essential to do the proactive work in preparing ourselves and our students for what it means to teach and learn in the age of AI. Rather than taking a reactive stance that requires us to deal with cheating that is hard to detect–and therefore can erode the trust between students and faculty that is essential for learning to flourish–let’s grapple with the AI beast head-on. Even if you have a firm belief that you will never allow its use in your courses, it is essential that you learn about GenAI’s capabilities and test your assignments and assessments to discover how vulnerable they are to its power. Only from an educated point of view can you know best how to manage teaching and learning in the age of AI. Let us help you! We’re here all summer and happy to help. And, when you are ready to integrate AI in teaching and learning, we’re here for that too. Look for workshops and resources coming this summer, or make an appointment with one of our team.

I would like to thank my team for adjusting magnificently to meeting faculty’s needs despite having to do so with a smaller team (we too have been dealing with the budgetary realities). I feel so lucky to have my wonderful team – couldn’t do it without them! 

Thank you for your hard work, your dedication, and your inspiring commitment to our students and each other. Well said, ChatGPT! In addition, dear colleagues, I am sending you warmest gratitude for your engagement with the CAT. Best wishes for a restful and enjoyable summer

Mindful Management of AI During Finals

by Dana Dawson

As we near the end of the semester, it’s important to carefully consider your plan of action should you suspect students have used generative AI in a manner that you explicitly prohibited. In past blog posts, we strongly encouraged faculty members to begin by meeting with the student in cases where you suspect unacceptable use of AI and to start with a conversation. However, in the case of final exams and projects, you may feel you don’t have time for that course of action. In this post, we offer some suggestions for how to prepare for and address AI use during the finals period.

Ensure Guidelines Are Clear

Review your final exams and final project instructions to determine whether you have clearly outlined where the use of generative AI is and is not allowed. Build guidelines into assignments as well as the syllabus to ensure students have it readily available. Have a conversation with your classes to ensure they understand the limitations of acceptable generative AI use and state the steps that will be taken if you suspect students have used generative AI (more on that below). 

Test Your Final Exams and Final Projects Using Generative AI

Run final exam questions or final project prompts through tools such as ChatGPT and Claude.AI and prompt the tools to take the exam or complete the project. Note that in ChatGPT, you can simply copy and paste the entire exam or project prompt and rubric into the tool and ask it to generate a response. Claude.AI allows you to upload a pdf and enter a prompt. If you find that the tools can successfully complete your exams or assignments, reconsider the questions and prompts. Can you link questions or project prompts to in-class work that will draw on students’ past experiences? Can you add reflective or metacognitive questions that are difficult to replicate using generative AI? See this EDvice Exchange blog post for assessment ideas that are less prone to AI use.

Be Wary of AI Detectors

It has been well-established that AI detectors are not reliably able to differentiate between human- and AI-written text. Assessments we conducted of Turnitin’s AI detector, and four other applications available for free online, show that these detectors are prone to false positives (identifying human-written text as generated by AI) and false negatives (identifying AI-written text as generated by humans). AI detectors should never be used as the sole basis for a judgment on whether a student has used AI; companies such as Turnitin acknowledge this, for example, saying in their own explanatory materials that detector predictions should be taken with a grain of salt and that the instructor must ultimately make their own interpretations. Notably, TurnItIn also indicates that a score of 20% or less AI-created should not be considered valid. As you assess AI detector reports, keep in mind that there are currently no completely reliable detectors of generative AI use in writing available to instructors.

Step on the Brakes

Confronting possible cheating is always stress-inducing. We see a block of text or a pattern of answers that seem unlikely to have been generated by a student and the stress response kicks in. This is not the optimal time to make a decision. Take a breath, step away. Consider factors that might be influencing your assessment of the student’s work or your willingness to accept the results of an AI detector. Talk to a colleague or a CAT consultant and carefully consider all factors before making a determination as to your course of action.

You Can Still Have a Conversation with Students

If you strongly suspect a student of using generative AI in a manner you have stated is not acceptable, ask the student to meet, by Zoom if they are already off campus. If they are not able to meet prior to the end of the grading period, issue an Incomplete for the course and do not grade the final exam or project until you have met with the student. 

Have a Back-Up Plan

If you speak with the student and they do not admit to using generative AI, have an actionable plan for how to proceed. Consider how you might replicate the element you suspect they used AI to complete. Can you conduct an oral exam? Can they write an essay or a reflective statement on their process of solving the exam question or completing the project in-person? To talk over your plan for considering possible AI use in these final weeks of the semester, don’t hesitate to reach out to schedule a consultation with a CAT specialist. 

Err On the Side of Caution

The suspicion that a student may be taking shortcuts can be upsetting and we are all struggling to manage course design and delivery in the age of AI but the risk of falsely accusing a student should be taken very seriously. A false accusation can derail a student’s entire educational trajectory and not only because of the possible impact on their GPA; more importantly, it can shake their trust in their faculty members, their experience with higher education and their motivation to continue, particularly where their sense of belonging is tenuous. Turnitin has acknowledged that their detector is more likely to generate a false positive in the case of English language learners or developing writers as some of the writing patterns more common among these populations are the same patterns AI detectors look for in identifying AI-generated text. We must exercise the utmost caution in accusing any student and be sure to give them the benefit of the doubt when engaging in these conversations. 

Plan for Next Semester

Finally, once finals are over and your grades are in, make an appointment with a CAT specialist to explore how to revise assignments that are particularly vulnerable to AI use. We can often avoid these problems in the future by revising our current assessments into ones that work better in the age of AI.

Faculty Adventures in the AI Learning Frontier: One Professor’s Take

by Jeff Rients, Ph.D.

Title card: Faculty Adventurers in the AI Learning Frontier

This week we’re happy to share a video featuring Michael L. Schirmer, who teaches the course Integrative Business Practices in the Fox School of Business. Michael shares his experiences with generative AI both with his students and as part of his personal scholarly practice.

Thank you so much for your insights, Michael!

If you’d like more guidance on exploring how to use AI tools in your class or assistance running your assignments through GenAI to better assess the value of using it, please visit our Faculty Guide to A.I. or book an appointment for a one-on-one consultation.

Faculty Adventures in the AI Learning Frontier: Assignments and Activities that Address Ethical Considerations of Generative AI Use

by Benjamin Brock, Ph.D and Dana Dawson, Ph.D

Title card: Faculty Adventurers in the AI Learning Frontier

In response to our fall 2023 survey on the use of generative AI (GenAI) in the classroom, we received a number of assignments and activities faculty members have designed to tackle the ethical issues raised by GenAI. Ethical concerns related to GenAI include such considerations as the implications for privacy when these tools are used, the possibility of over-reliance on GenAI for analytics and decision making, and exposure to inaccurate or biased information (Brown & Klein, 2020; Masters, 2023; Memarian & Doleck, 2023). The following activities and assignments equip students with the capacity to critically evaluate when and how it is appropriate to use GenAI tools and to protect themselves against possible risks of AI use.

Sherri Hope Culver, Media Studies and Production faculty member and Director of the Center for Media and Information Literacy (CMIL) at Temple University, asks students in her GenEd course, Media in a Hyper-Mediated World, to complete a reflection on the implications of AI use. She first asks them to listen to an episode of the podcast Hard Fork centered on data privacy and image manipulation and to read the Wired article “The Call to Halt ‘Dangerous’ AI Research Ignores a Simple Truth” (Luccione, 2023). Students are then instructed to write a 300-word reflection referencing the assigned material that addresses both concerns they have about use of AI and ways in which it could make their lives or society better. Professor Culver provides the following prompts to help students’ thinking:

  • What does critical thinking mean in a tech-centric, AI world?    
  • How might AI affect your free will?    
  • How might AI affect your concerns about privacy or surveillance?    
  • How should we prepare ourselves for an increasingly AI world?    
  • How might AI influence the notion of a public good?   
  • How might AI influence K-12 education?    
  • How might AI influence family life?    
  • What worries you about AI?    
  • What excites you about AI?    
  • What is our responsibility as media creators when we use AI?    
  • It has been said that AI will make life more “fast, free and frictionless.” Should everything first be “fast, free and frictionless”? Should that be the aim?
  • Is AI the end of truth?

In a dynamic, interactive, reflection-oriented honors course aimed at exploring the four pillars of Temple’s Honors Program (inclusive community, intellectual curiosity, integrity in leadership, and social courage), Dr. Amanda Neuber, Director of the Honors Program, is using AI as the discussion anchor for their unit on “integrity in leadership.” By way of multiple media modalities, students delve into the ethical and unethical uses of AI in academia. Students are asked to read “How to Use ChatGPT and Still Be a Good Person” and watch a related video exploring the meaning of integrity. Students then discuss whether or not AI can be used with integrity, how academic culture might frame one’s decision to use AI, and the “peaks and pitfalls” of AI use. Beyond the many important conversations focused on AI itself, the technology is used as a reference point as to what it means to lead with integrity and how to promote said quality in teams and organizations.

In another interactive, thought-based classroom initiative, mechanical engineer Dr. Philip Dames is bringing ethics and AI to Temple’s College of Engineering. Having reimagined for a modern era the “trolley problem” philosophical exercise in which one is faced with an ethical dilemma, students in Dr. Dames’ class consider having AI make decisions using autonomous cars as the basis for deliberation. They are prompted to think about how a vehicle should be programmed to respond to different scenarios by using examples from MIT Media Lab’s Moral Machine website. Students then reflect upon their scenario-based activities and experiences and engage in prompt-guided written reflection. Prompts include questions such as: 

  • How does the ownership model of autonomous vehicles affect how they should behave? For example, does it make a difference if a vehicle is owned by a single private citizen vs. publicly owned by the city and hired by individuals? 
  • What surprised you about the aggregated responses from different people shown to you at the end of the exercise? 
  • Are there other factors that you feel are important but were not considered in Moral Machine?

In this way, students not only explore elements to consider when designing autonomous vehicles, but make concrete what was once only abstract via critical thinking and hands-on engagement.

If you’d like more guidance on exploring how to use AI tools in your class, please visit our Faculty Guide to A.I. and/or book an appointment for a one-on-one consultation.

Brown, M., & Klein, C. (2020). Whose data? Which rights? Whose power? A policy discourse analysis of student privacy policy documents. The Journal of Higher Education, 91(7), 1149–1178. https://doi.org/10.1080/00221546.2020.1770045  

Masters, K. (2023). Ethical use of artificial intelligence in health professions education: AMEE Guide No. 158. Medical Teacher, 45(6), 574–584. https://doi.org/10.1080/0142159X.2023.2186203  

Memarian, B., & Doleck, T. (2023). Fairness, accountability, transparency, and ethics (FATE) in artificial intelligence (AI) and higher education: A systematic review. Computers and Education: Artificial Intelligence, 5 (2023), 1-12. https://doi.org/10.1016/j.caeai.2023.100152