Teaching Programming With AI Is Now an Evidence-Based Decision
Most conversations about generative AI in programming courses still run on instinct. One camp wants to ban the tools to protect the fundamentals. The other wants to adopt them quickly so students are ready for an industry that has already changed. Both positions get argued from conviction, and both skip the step that should come first: looking at what the research says.
That step is now possible. In the past two years, computing education researchers have produced enough studies to move the question from opinion to evidence. The findings do not hand a clean win to either camp. They support bringing AI into the programming classroom, and they show that doing it badly can widen the gap between students who are already succeeding and those who are struggling. The case for integration and the case for careful design turn out to be the same case.
What the tools are good at
Start with the capability question, because for the types of problems we usually assign, it is no longer in serious dispute. Current large language models solve the kinds of problems we assign in introductory courses. A review of the field by Paul Denny and colleagues documents how code-generation tools went from a curiosity to a standard part of the programming workflow in a matter of months, and how readily they handle the small, well-specified problems that fill most CS1 assignments. The exam evidence is concrete. When James Finnie-Ansley and colleagues ran OpenAI’s Codex through two CS1 tests in 2022, its answers ranked seventeenth in a class of seventy-one, inside the top quartile, and that was an early model. Newer ones have only widened the margin.
The industry signal points the same direction. In April 2025, Microsoft’s CEO said as much as 30% of the company’s code was being written by AI. That figure is a remark on a conference stage, not a measured result, and “written by AI” has no agreed definition, so it deserves caution. As a marker of where professional practice is heading, though, it is hard to ignore. If we want graduates who can work the way their future colleagues already work, ignoring the tools is its own kind of risk.
The second strength is tutoring. A first-year case study by Liam Stienstra and colleagues used a ChatGPT-based tutor to teach a new topic, then tested students without assistance. The tool supported learning, and most students preferred the AI tutor to a human one. Their reasons are worth considering: it was always available, and it did not make them feel judged for asking a basic question. Purpose-built tutors push this further. Tools like CodeHelp, from Mark Liffiton and colleagues, and CodeAid, from Majeed Kazemitabaar and colleagues, wrap the model in guardrails so it nudges a student toward an answer instead of handing one over. Work by Harsh Kumar and colleagues, run across a classroom study and a larger controlled experiment, found that the structure an instructor puts around the tool changes how students use it. Guidance that prompted students to attempt a problem before turning to the model, or to reflect on their own thinking, cut down on aimless queries and on the habit of pasting an assignment straight into the chatbot verbatim without attempting it first.
There is also early evidence that the help does not vanish the moment the tool is taken away. In a controlled study, Kazemitabaar and colleagues gave sixty-nine beginners, ages ten to seventeen, a sequence of forty-five Python tasks. Half had an AI code generator during the authoring work. That group completed more tasks and scored higher on the authoring work. The natural worry is that they would fall apart once the tool was gone. They did not. On the modification tasks they had to do by hand, they performed no worse than the students who never had access. On a retention test a week later they came out slightly ahead, though the difference was not statistically significant. Encouraging and modest at once, which is the honest way to hold it.
So far this is the authentic optimistic story. It is also incomplete.
The catch is an equity problem
Learning with AI is not automatic, and the students it fails are the ones we should worry about most.
The clearest evidence comes from James Prather and colleagues, in a study with the apt title “The Widening Gap.” Using eye-tracking and think-aloud methods with novice programmers, they found that every metacognitive difficulty students had before AI is still present, and that new ones have appeared. The pattern that is concerning: students with higher grades and stronger self-efficacy used the tools to accelerate, while weaker students were often slowed down and left with what the authors call an illusion of competence. The work finished but the understanding did not. The same tool that helps a confident student move faster can help a struggling student fail without noticing.
Part of the reason can be found in a paper by Lev Tankelevitch and colleagues on the metacognitive demands of generative AI. Working with these systems asks the user to do things that are cognitively hard: state a goal precisely, break it into parts, judge whether the output is any good, and adjust course when it is not. The authors compare it to a manager delegating to a team. That is a demanding skill, and it is not one novices arrive with. Studies of how beginners actually use these tools back this up. A close look by James Prather and colleagues at novices working with Copilot found students leaving out details that matter, like input and output types, and, when a prompt failed, changing a few words in the prompt rather than supplying the missing requirements.
This is where the equity lens is needed. A tool that rewards students who already plan well, monitor their own understanding, and know what they do not know will tend to reward the students who were going to do fine anyway. Left unaddressed AI creates a larger equity gap.
What evidence-based design looks like
The same body of research points to what helps, and the through-line is consistent: slow students down enough to make them think, and build the supports that strong students supply for themselves.
Majeed Kazemitabaar and colleagues tested this directly. They compared a baseline that simply showed students AI-generated code with an explanation against seven techniques that required some engagement with the code before or after it appeared. The technique that worked best, called Lead-and-Reveal, brought what students thought they had learned closest to what they actually learned, without adding cognitive load. The effects were modest and the samples small, which the authors are candid about, but the direction matches everything else in the literature. Passive acceptance of generated code produces the illusion of learning. Forcing a small act of reasoning interrupts it.
Other design moves follow from the same logic. Scaffold the tool rather than handing it over, as Kumar’s guidance strategies did. Teach prompting as an explicit skill, since novices do not pick it up on their own. Move assessment toward process and away from the take-home artifact that a model can now generate in seconds. And give upper-level students structured practice with the messier task they will actually face at work: a study by Anshul Shah, Leo Porter, and colleagues taught students to use GitHub Copilot inside a large codebase and found a pattern they call one-shot prompting, where students ask the tool to build an entire feature at once and then spend their effort debugging what comes back. That is a teachable habit, and naming it is the first step to correcting it.
There is a quieter finding that reframes the whole debate. A 2025 replication study by Rita Garcia and Michelle Craig revisited a 2004 survey of what CS1 instructors teach and find hard to teach. Twenty years on, the basic concepts are largely unchanged, recursion is still the hardest thing to teach, and the newest challenge instructors name is teaching students to think and plan before they code. That difficulty predates generative AI. The shift toward planning, decomposition, and self-monitoring, the very skills AI makes more important, was already underway. The course it was happening in was already struggling: introductory programming has carried high failure rates for decades, documented by Jens Bennedsen and Michael Caspersen and revisited by Christopher Watson and Frederick Li, and the long line of work on Soloway’s rainfall problem, surveyed by Kathi Fisler, showed that many students finish CS1 unable to write a basic loop correctly. AI did not break the introductory course. It exposed problems the course already had.
You do not have to figure this out alone
The research is scattered across conferences and journals, which makes it hard for any individual instructor to assemble. That gap is part of what the Consortium for Generative AI in CS Education https://www.teachcswithai.org/about exists to close. Housed at UC San Diego’s Center for Research on Education, Assessment, and Teaching Excellence and led by Leo Porter, Daniel Zingaro, and Beth Simon, the consortium gathers course materials, research summaries, and a community of educators working through the same questions. Its partner network runs from professional bodies like ACM and SIGCSE to organizations focused on access and inclusion, which matters given what the research shows about who gets left behind. The point of the consortium is not to sell a position. It is to make the evidence usable, so that a faculty member redesigning a course in the summer is not starting from a blank page. About half of the articles I used for this article can be found in the consortium’s resources section.
Two of the consortium leads, Porter and Zingaro, have also written the classroom-ready version. Their book, Learn AI-Assisted Python Programming, reorders the introductory course around the skills the research elevates. They argue that the skills to write good software are evolving: problem decomposition, writing a specification, reading code, and testing matter more than they used to, while memorizing syntax and library details matters less. Two of the chapters teach students to read code, on the logic that if the assistant writes the code, the student’s job is to judge whether the assistant has done what the student intended. A full chapter covers testing and prompt engineering. Top-down design, the practice of breaking a large problem into smaller ones, runs across several chapters rather than appearing once. Debugging gets its own deep dive, a skill that think-aloud studies by Jacqueline Whalley and colleagues show novices tend to approach unsystematically. Porter and Zingaro grant that a future professional engineer should eventually learn to write code from scratch, but they argue it no longer makes sense as the starting place for most learners. That is a concrete answer to the camp that wants to ban the tools to protect the fundamentals: they are just sequenced differently and weighted toward the skills that matter most now, like reading code, testing, and breaking problems down, rather than writing syntax from memory. The book is available to the IU community through IUCAT on the Bloomington and Indianapolis campuses, in digital ebook (https://iucat.iu.edu/catalog/20590350) and streaming audio (https://iucat.iu.edu/catalog/21587005) formats.
What about large courses?
Class size does not change the goals, but it splits these design moves into two groups. Some work just as well with three hundred students as with thirty and cost the instructor no extra effort. Requiring an attempt before a student reaches for the tutor is a policy built into the assignment or the tool, not labor spent per student. Kumar’s guidance strategies were tested as exactly this kind of scalable intervention. The cognitive-engagement step is the same: Kazemitabaar’s Lead-and-Reveal shipped as a reusable task builder, so the prediction step runs for every student without the instructor touching it. The AI tutor itself has its strongest case here. Kumar frames LLM tutors as a response to growing class sizes where teacher presence is thin, CodeHelp was built for scalable support in large classes, and Stienstra’s students valued the tutor because it was always available, which a human teaching assistant in a 400-person course cannot be.
The moves that resist scale are both forms of assessment. Reading hundreds of process artifacts or grading prompts by hand does not work past a certain size. The realistic versions are to make the process artifact auto-gradeable or peer-reviewable, push it into teaching-assistant-led lab sections, assess prompts by outcome rather than by reading each one, and reserve the labor-heavy formats like oral checks for a rotating sample. The teaching of these skills scales fine, since a lecture is a lecture. The grading is what has to be redesigned.
Underneath this is the equity point. In a small class you can see who is drifting. In a large one, the student coasting on an illusion of competence is invisible, and that is the student Prather’s widening-gap study identifies as most at risk. The large-course adaptation is a triage logic: let the tools that scale support the median student, and aim scarce human attention, teaching-assistant hours, oral checks, early-alert outreach, at the students the tools are most likely to fail. Spreading human help evenly is the thing a big course cannot afford and the thing struggling students most need.
The opportunity
The strongest argument for bringing AI into programming education is not that the tools are impressive, though they are. It is that they force a reckoning with weaknesses the field has tolerated for decades: introductory courses where many students do not learn what we think they learn, and a persistent distance between what we teach and what graduates need. We can treat that reckoning as a threat to manage or as a chance to fix the course. The evidence says the fix is within reach, and that it depends less on the tools than on how we design around them.
A few things to consider trying next term:
Replace one take-home assignment with an in-class task that assesses process, not just the finished program.
Before students use an AI tutor, require them to attempt the problem and write down what they think the answer should be.
Teach one short lesson on writing a specification for a model, including input and output types, and grade a prompt the way you would grade code.
Pick a single cognitive-engagement move, such as having students predict what generated code will do before running it, and use it consistently for a unit.
References
Denny, P., Prather, J., Becker, B. A., Finnie-Ansley, J., Hellas, A., Leinonen, J., Luxton-Reilly, A., Reeves, B. N., Santos, E. A., & Sarsa, S. (2024). Computing Education in the Era of Generative AI. Communications of the ACM. https://doi.org/10.1145/3624720
Finnie-Ansley, J., Denny, P., Becker, B. A., Luxton-Reilly, A., & Prather, J. (2022). The Robots Are Coming: Exploring the Implications of OpenAI Codex on Introductory Programming. Australasian Computing Education Conference (ACE ’22). https://doi.org/10.1145/3511861.3511863
Kazemitabaar, M., Chow, J., Ma, C. K. T., Ericson, B. J., Weintrop, D., & Grossman, T. (2023). Studying the Effect of AI Code Generators on Supporting Novice Learners in Introductory Programming. CHI ’23. https://doi.org/10.1145/3544548.3580919
Prather, J., Reeves, B. N., Denny, P., Becker, B. A., Leinonen, J., Luxton-Reilly, A., Powell, G., Finnie-Ansley, J., & Santos, E. A. (2023). “It’s Weird That it Knows What I Want”: Usability and Interactions with Copilot for Novice Programmers. ACM Transactions on Computer-Human Interaction. arXiv:2304.02491
Liffiton, M., Sheese, B., Savelka, J., & Denny, P. (2023). CodeHelp: Using Large Language Models with Guardrails for Scalable Support in Programming Classes. Koli Calling ’23. arXiv:2308.06921
Kazemitabaar, M., Ye, R., Wang, X., Henley, A. Z., Denny, P., Craig, M., & Grossman, T. (2024). CodeAid: Evaluating a Classroom Deployment of an LLM-based Programming Assistant that Balances Student and Educator Needs. CHI ’24. arXiv:2401.11314
Prather, J., Reeves, B. N., Leinonen, J., MacNeil, S., Randrianasolo, A. S., Becker, B. A., Kimmel, B., Wright, J., & Briggs, B. (2024). The Widening Gap: The Benefits and Harms of Generative AI for Novice Programmers. ICER. https://doi.org/10.1145/3632620.3671116
Tankelevitch, L., Kewenig, V., Simkute, A., Scott, A. E., Sarkar, A., Sellen, A., & Rintel, S. (2024). The Metacognitive Demands and Opportunities of Generative AI. CHI ’24. https://doi.org/10.1145/3613904.3642902
Kumar, H., Musabirov, I., Reza, M., Shi, J., Wang, X., Williams, J. J., Kuzminykh, A., & Liut, M. (2024). Guiding Students in Using LLMs in Supported Learning Environments: Effects on Interaction Dynamics, Learner Performance, Confidence, and Trust. CSCW. https://doi.org/10.1145/3687038
Kazemitabaar, M., Huang, O., Suh, S., Henley, A. Z., & Grossman, T. (2025). Exploring the Design Space of Cognitive Engagement Techniques with AI-Generated Code for Enhanced Learning. IUI ’25. https://doi.org/10.1145/3708359.3712104
Stienstra, L., Mohamed, A., & Mohamed, M. (2025). Exploring GenAI as a Tutoring Tool: A Case Study in First-Year Computer Programming. ITiCSE ’25. https://doi.org/10.1145/3724363.3729060
Shah, A., Chernova, A., Tomson, E., Porter, L., Griswold, W. G., & Soosai Raj, A. G. (2025). Students’ Use of GitHub Copilot for Working with Large Code Bases. SIGCSE ’25. https://doi.org/10.1145/3641554.3701800
Garcia, R., & Craig, M. (2025). 20 Years Later: A Replication Study on Teaching CS1 Concepts. ACM Transactions on Computing Education. https://doi.org/10.1145/3730405
Whalley, J., Settle, A., & Luxton-Reilly, A. (2023). A Think-Aloud Study of Novice Debugging. ACM Transactions on Computing Education. https://doi.org/10.1145/3589004
Fisler, K. (2014). The Recurring Rainfall Problem. ICER ’14. https://doi.org/10.1145/2632320.2632346
Watson, C., & Li, F. W. B. (2014). Failure Rates in Introductory Programming Revisited. ITiCSE ’14. https://doi.org/10.1145/2591708.2591749
Bennedsen, J., & Caspersen, M. E. (2007). Failure Rates in Introductory Programming. ACM SIGCSE Bulletin. https://doi.org/10.1145/1272848.1272879
Porter, L., & Zingaro, D. Learn AI-Assisted Python Programming, Second Edition: With GitHub Copilot and ChatGPT. Manning (Function). Kindle Edition. Available via IUCAT, Bloomington and Indianapolis campuses: ebook https://iucat.iu.edu/catalog/20590350, streaming audio https://iucat.iu.edu/catalog/21587005
Novet, J., & Vanian, J. (2025, April 29). Satya Nadella says as much as 30% of Microsoft code is written by AI. CNBC.
Consortium for Generative AI in CS Education, UC San Diego CREATE. Faculty leads: Leo Porter, Daniel Zingaro, Beth Simon. https://www.teachcswithai.org/
Porter, L. Research on GenAI and Learning To Program [Video]. YouTube. https://www.youtube.com/watch?v=faY5gDTlvs0
Beyond the Hype: A Practical GenAI Resource Guide for Faculty in Technical Disciplines
As faculty that teach technical disciplines, you are in a unique position. You aren’t just figuring out how to use Generative AI; you are teaching the students who will build, deploy, and critically evaluate these tools for years to come.
The challenge is twofold:
How can you leverage AI to improve your own teaching (e.g., create coding examples, debug assignments, or design better projects)?
How can you effectively integrate AI into your curriculum as a core competency (e.g., teach prompt engineering, model limitations, and AI ethics)?
The internet is flooded with AI resources, and it’s impossible to sift through them all. This post is a practical, curated guide to help you find the most useful resources for your courses without the noise.
Start with IU: Key Local Resources
Before diving into the wider web, start with the excellent resources available directly from IU. These provide the foundational context and policies for our community.
Generative AI 101 Faculty Resources
Description: An overview of the GenAI 101 Course available to all at IU. Also includes a syllabus insert that can be used to promote the course to students.
Kelley School of Business “AI Playbook”
Description: A “living guide” developed by the Kelley School for faculty on the use of generative AI in teaching, grading, and research. It outlines shared values and emphasizes that faculty expertise remains central.
When to use: When you want faculty-facing guidance on when and how to use generative AI in assessments, course design, and feedback workflows.
A Quick Starting Point: Three Actionable Resources
If you want to branch out, here are three high-value resources to review in 10 minutes or less.
For Your Curriculum: Teach CS with AI: Resource Hub for Computer Science Educators
What it is: A hub specifically for integrating AI into CS courses. It includes lesson plans, project ideas, and pedagogical strategies for teaching AI in computing.
When to use: When you’re not just using AI, but actively teaching AI concepts, ethics, or applications within a CS or Informatics course.
For Your Pedagogy: Harvard University:“Teaching with Gen-AI” resources
What it is: High-level guidance from Harvard on course design, with excellent case studies and strategies for handling risks like hallucinations and superficial reasoning.
When to use: Use this before the semester starts. It’s perfect for designing your syllabus, setting AI policies, and building responsible use guidelines into your course from day one.
For Your Students (and You): AI for Education: “Effective Prompting for Educators”
What it is: A focused guide on how to write better prompts. It includes frameworks (like the “5 S Framework”) that are perfect for teaching students a structured approach to “prompt engineering.”
When to use: When you want to move students beyond simple “ask-and-receive” and teach them how to partner with AI to get better, more reliable, and more complex results.
The Deep Dive: A Curated Resource Library
For those with more time, here is a more comprehensive list organized by task.
1. How to Use AI in Your Classroom (Pedagogy & Assignments)
Derek Bok Center for Teaching and Learning (Harvard) : “Teaching in the Age of AI”
Description: A collection highlighting definitions, pedagogical questions, and course design implications.
When to use: When you want concrete assignment ideas or sample activities to see how others have used AI in their teaching.
One Useful Thing: Blog post “15 Times to Use AI and 5 Not To”
Description: Easy-to-digest guidance on where AI is helpful in learning (quantity-based tasks, translation) and where it isn’t (deep learning, high-accuracy needs).
When to use: When you are deciding which parts of your course are appropriate for AI and which should remain human-only.
Syllabi Policies for AI Generative Tools:
Description: A collection of sample policies from other educators to help you develop your own.
When to use: When you are writing your syllabus and need clear language about how AI tools are (or aren’t) allowed in student work.
2. Helping Students (and You) Get Better at Prompting
AI for Education: Prompt Library
Description: A comprehensive, searchable collection of ready-to-use prompts and templates specifically for educators.
When to use: When you need quick, plug-and-play prompt templates for lesson plans, student tasks, or administrative work.
More Useful Things — Prompt Repository for Educators
Description: A repository of prompts for instructor aids and student exercises, curated by researchers Ethan and Lilach Mollick.
When to use: When you want tested, inspiring prompt sets, especially for idea generation or in-class activities.
Description: Anthropic’s (maker of Claude) public library of optimized prompts for business, creative, and general tasks.
When to use: When you want to show students (or yourself) “what good prompting looks like” from an industry leader.
3. How to Teach AI in Your CS/InF Courses (Curriculum & Literacy)
Teach CS with AI: Resource Hub for Computer Science Educators
Description: A hub dedicated to integrating AI topics, tools, and teaching strategies in CS courses.
When to use: Use when teaching a CS course and you want to integrate AI content (topics, labs, projects) directly.
metaLAB at Harvard: The AI Pedagogy Project / AI Guide
Description: A curated site with assignments and projects to integrate AI in pedagogical practice, focused on critical thinking.
When to use: When you are designing a module on AI literacy, critical AI thinking, or assessing students’ interaction with AI tools.
Ideeas Lab: Teaching & AI resources
Description: A resource hub with teaching materials and tools, particularly aimed at engineering and technical fields.
When to use: When you want resources specifically tailored for engineering domains that integrate AI in assignments.
AI for Education: “Generative AI Critical Analysis Activities
Description: Classroom activities to help students critically examine AI outputs, ethics, and limitations.
When to use: When you want to design modules around AI ethics or have students evaluate AI rather than simply use it.
4. Taking it Further: Building Your Own AI Tools
HBSP: “How to Create Custom AI Chatbots that Enrich Your Classroom”
Description: A walk-through article for educators on designing custom AI chatbots for classroom engagement.
When to use: Use when planning to design or integrate a custom AI chatbot (e.g., as a teaching assistant, for Q&A) in your class.
OpenAI: “A Practical Guide to Building Agents” (PDF Business/Technical Guide)
Description: A detailed technical guide from OpenAI on building AI agents, including design, steps, and deployment.
When to use: Use when you or your students are planning to build or deploy an AI agent as part of a capstone or advanced project.
5. Professional Development & Staying Current
IBM Skills Build for Educators: College Educators resources
Description: A professional development site offering modules and training materials to build AI fluency and integrate digital skills into teaching.
When to use: When you want a structured PD path for yourself or want to build a course around AI literacy and workforce readiness.
University of Maine: LearnWithAI initiative
Description: A practical, “how-to” oriented site for faculty on integrating AI into courses.
When to use: Use when you want a site focused on faculty development and practical course integration.
Future-Cymbal Notion Page: Shared collection of AI-Teaching Resources
Description: A collaboratively curated Notion page of ideas, links, frameworks on AI in education; less “formal guide,” more open resource aggregation
When to use: Use when you want to browse a broad, ever-updating set of ideas rather than a polished handbook.
Description: It collects a wide variety of resources for educators around generative AI in the classroom — such as sample syllabus statements, institutional policy templates, teaching ideas, and faculty development materials.
When to use: When you are designing or revising your course syllabus and need clear language about how you will (or won’t) allow AI tools in student work.
Newsletters for Staying Current:
The Rundown -Daily newsletter summarizing AI news across research, policy, and industry.
The Neuron – Broad coverage of emerging AI trends and commentary, often with education-adjacent insights.
The Batch – Weekly deep dives into AI research, tools, and development—ideal for those following the tech side.
The Algorithmic Bridge | Alberto Romero – Thoughtful essays analyzing AI’s social, ethical, and educational impact.
Everyday AI Newsletter – Daily newsletter (and accompanying podcast) aimed at making AI accessible to “everyday people” whether educators, professionals, or non-tech specialists.
Conclusion: Start Small, Start Now
You don’t need to redesign your entire curriculum overnight. The best approach is to start small.
Pick one thing to try this month. It could be using a prompt library to help you write a coding assignment, adapting a syllabus policy, or introducing one critical analysis activity in a senior seminar. By experimenting now, you’ll be better prepared to lead your students in this new, AI-driven landscape.
Did I miss a great resource? Leave a comment and let me know!
Teaching for Integrity in the Age of AI: From Compliance to Culture
Inspired by Chapter 2 of The Opposite of Cheating: Teaching for Integrity in the Age of AI by Tricia Bertram Gallant and David Rettinger
Academic integrity is not a checklist or compliance form. It is a living culture shaped by what we model, how we design, and the conversations we hold with our students. Gallant and Rettinger remind us that integrity is cultivated through transparency, design, and dialogue, not surveillance or punishment. The real challenge now is how to teach integrity in an age where AI is everywhere.
The U.S. Department of Education’s 2023 report, “Artificial Intelligence and the Future of Teaching and Learning”, encourages educators to treat AI as a design opportunity for advancing human-centered learning, not a threat to academic honesty. Recent data highlight the urgency of this work. According to the Higher Education Policy Institute’s 2025 Student AI Survey, 92% of undergraduates report using generative AI tools, up from 66% the year before. A Guardian report found that AI-related misconduct cases have tripled since 2023. The takeaway is clear: integrity education has to evolve alongside AI literacy.
2025 Snapshot: AI & Academic Integrity
Use and Attitudes
Over 85% of undergraduates use GenAI tools (Inside Higher Ed, HEPI 2025)
61% of students want clear, course-level AI policies
33% of students were concerned about being accused of plagiarism or cheating (Campus Technology)
While 45% believe using AI for editing is “acceptable academic support”
Institutional Responses
A survey found that less than 40% of higher education institutions have formal AI acceptable use policies. (EDUCAUSE 2025)
AI-related misconduct tripled in the UK (The Guardian 2025)
76% of universities are redesigning assessments for AI contexts (UCL Education AI Survey 2025)
Faculty Trends
A significant gap exists between student and faculty adoption: only 61% of faculty report using AI in teaching, and of those, a large majority (88%) do so minimally (ASEE AI Training 2025 led by Drs. Adita Jori and Andrew Patz).
82% of instructors use GenAI for feedback or rubric design (EDUCAUSE 2025 AI Landscape Study)
Detection tools now use watermarking and metadata tracing, but false positives remain a major concern (arXiv 2025)
Model Integrity
Students notice how we work. They learn from the way we check our sources, document decisions, and acknowledge mistakes. Modeling integrity starts with transparency.
As the EDUCAUSE 2025 AI Landscape Study notes, many universities are investing in training that helps faculty engage AI responsibly. Modeling integrity now means showing how to use AI intentionally, not avoid it.
This aligns with findings from Gu and Yan’s 2025 meta-analysis, which showed that students benefit most when teachers scaffold AI use and talk openly about it. When instructors frame AI as a learning partner, not a shortcut, students develop stronger judgment and accountability.
Make Integrity Explicit
Integrity should show up as often in our discussions as it does in our policies. When we talk about it before projects, during collaborations, and after challenges, students begin to see ethics as part of the learning process.
Tricia Bertram Gallant and David Rettinger emphasize that ethical behavior thrives when it’s designed into the experience. Singer-Freeman, Verbeke, and Barre (2025) found that students across all academic levels want clear guidance on what’s acceptable AI use. If we make expectations explicit, we replace anxiety with understanding.
A recent MDPI review on Generative AI and Academic Ethics reinforces this point, noting that while GenAI can enhance engagement and efficiency, it also increases risks to originality and ethical reasoning.
Use Clear, Simple Language (The Social Institute)
Students need to understand A.I. policies to be able to follow them. That means avoiding jargon and overly technical language.
Instead of: “A.I. assistance must align with established academic integrity principles.”
Say: “You may use A.I. for brainstorming ideas, but not for writing entire sections of code or essays.”
Establish consistent Rules Across Departments or Schools
One of the biggest sources of confusion is inconsistent enforcement when it comes to A.I. rules. Departments or schools can develop a universal A.I. guidelines that applies to all instructors, rather than allowing individual educators to set conflicting rules. Over half of students (58%) report that their school or program has a policy, but a substantial number (28%) say it differs, with some courses or professors having a policy and some not (Forbes 2025). Consider creating an instructor handbook outlining departmental or school-wide A.I. best practices to make sure they are consistently communicated to students.
Frame Integrity Positively
Instead of framing integrity around rules, frame it around growth. Students respond better when they see ethical choices as part of their professional development.
A Packback editorial on academic integrity in 2025 argues that punitive detection systems often erode trust and discourage learning. When faculty shift from surveillance to conversation, integrity becomes something students take ownership of, not something they fear.
Clarify Expectations
Ambiguity creates rationalization. In the age of AI, clarity is an act of fairness.
The National Centre for AI’s 2025 student study found that first-year students, in particular, feel confused about when and how AI use is acceptable. Faculty can address this by defining boundaries early and discussing examples. Transparency about tools, citations, and documentation helps students learn discernment.
Research from arXiv’s 2025 watermarking study adds that while detection tools are improving, they still make errors. Open conversations about what these systems can and cannot do build trust and understanding. Institutions like MIT and Duke University (22 minute mark) provide sample policy language for faculty to adapt. These statements define what “appropriate help” means and require students to cite AI contributions when used. Clarity transforms anxiety into accountability.
Normalize Conversations About Ethics
Ethics belongs in everyday learning. Conversations about bias, authorship, and data use should happen alongside technical instruction.
A 2025 study on synthetic media ethics found that students value open discussions about deepfakes and misinformation but often lack the frameworks to evaluate them. Integrating these discussions into our teaching helps students connect ethics to both academic and professional practice.
Use the Syllabus as a Moral Document
The syllabus sets the tone for integrity. Transparent grading policies, clear AI statements, and flexible revision options communicate fairness and care.
Universities are redesigning their syllabi and assessments to support “authentic learning” instead of reactive policing. The University of Melbourne’s Assured Learning model and the UCL Education AI Initiativeare leading examples, focusing on oral exams, reflective portfolios, and transparent assessment design.
Respond to Misconduct Constructively
When integrity violations occur, they can become moments for growth. Reflection, accountability, and dialogue teach more than punishment ever could.
The Packback 2025 Integrity Report encourages “growth-oriented remediation,” noting that many flagged cases stem from confusion, not intention. At Indiana University, we can uphold policy while still approaching each case as a learning opportunity.
Building a Culture of Integrity
Integrity thrives when it’s shared across the institution. Faculty, staff, and students each play a role.
The University of New South Wales’ 2025 partnership with OpenAI illustrates this shift: giving staff controlled access to ChatGPT within a responsible use framework. When universities model integrity through their own practices, students learn that ethics is not a barrier to innovation—it’s the framework that sustains it.
Final Thought
Teaching for integrity in the age of AI is about creating conditions where honesty becomes the natural choice. When we model transparency, design for trust, and engage in open dialogue, we teach more than content—we teach character.
As Amanda McKenzie, Director of Academic Integrity at the University of Waterloo, Canada, shares, “Integrity is not the opposite of cheating. It’s the presence of purpose.” When that purpose runs through our teaching, policies, and partnerships, we do more than protect academic standards. We prepare students to lead with integrity in a world increasingly shaped by AI.
Building a Framework for Academia-Industry Partnerships and AI Teaching and Learning Podcasts
In March, I shared the an overview of the “Practitioner to Professor (P2P)‘ survey that the CRA-Education / CRA-Industry working group analyzed. They recently released a report titled Breadth of Practices in Academia-Industry Relationships which explores a range of engagement models from research partnerships and personnel exchanges to master agreements and regional innovation ecosystems.
Key Findings and Observations
The report organizes its findings from the workshop into three categories: observations, barriers, and common solutions:
Observations A major theme was the critical need to embed ethical training into AI and computing curricula through both standalone courses and integrated assignments. It was noted that while academia is best suited to drive curriculum development, input from industry is essential to ensure the content remains relevant to real-world applications.
Barriers Key barriers to successful collaboration were identified, including cultural differences and misconceptions between academic and industry partners. For instance, industry’s focus on near-term goals can clash with academia’s long-term vision. A significant practical barrier is the prohibitive cost of cloud and GPU hardware, which limits students’ experience with cloud and AI development tools.
Common Solutions Effective solutions include the fluid movement of personnel between organizations through internships, co-ops, sabbaticals, and dual appointments. Streamlined master agreements at the institutional level also help facilitate research collaborations by reducing administrative friction.
Strategies for Research Collaboration
The report outlines a multi-level approach to enhancing research partnerships:
Individuals Faculty and industry researchers can initiate relationships through internal seed grants, sabbaticals in industry, dual appointments, and by serving on industry advisory boards.
Departments Departmental leaders can foster collaboration by strategically matching faculty expertise with industry needs, offering administrative support, and building a strong departmental brand with local industry.
University Leadership Senior leaders can address systemic barriers by creating a unified, institution-wide strategy, developing flexible funding models, and implementing master agreements to streamline partnerships.
Regional Ecosystems The report emphasizes the importance of universities partnering with local industries and startups to build thriving regional innovation ecosystems, which can drive economic development and secure government support.
Education and Workforce Development
With the rise of generative AI, the report highlights an urgent need for universities and industry to partner on education.
Curriculum Adaptation Computing curricula need to be updated to include foundational concepts in DevOps and scalable systems, which are often not part of the core curriculum. While AI literacy is essential, the report suggests a balance, with 80% of instruction remaining focused on core computer science skills. Ethical reasoning should be integrated throughout the curriculum, not just in a single course.
Workforce Programs To meet industry demands for job-ready graduates, the report advocates for university-industry partnerships in co-op programs, internships, and capstone projects. It also points to the need for universities to offer flexible programs like certificates and online courses to help upskill and reskill the existing workforce.
Recommendations
The report concludes with five main recommendations for universities, industry, and government:
Enhance research impact by combining academia’s long-term vision with real-world problems from industry. This can be achieved by embedding faculty in industry and industry researchers in universities.
Leverage the convening power of universities to build partnerships that benefit the wider community, using mechanisms like industrial advisory boards and research institutes.
Accelerate workforce development by aligning university programs with regional innovation ecosystems and having industry invest in talent through fellowships and internships.
Deliver industry-relevant curricula grounded in core computing principles, and collaborate with industry experts to co-design courses in high-demand areas like AI and cloud computing.
Establish new incentives and metrics to recognize and reward faculty for their contributions to industry partnerships in promotion and tenure evaluations.
AI Teaching and Learning Podcasts:What If College Teaching Was Redesigned With AI In Mind?
https://learningcurve.fm/episodes/what-if-college-teaching-was-redesigned-with-ai-in-mind
A former university president is trying to reimagine college teaching with AI in mind, and this year he released an unusual video that provides a kind of artist’s sketch of what that could look like. For this episode, I talk through the video with that leader, Paul LeBlanc, and get some reaction to the model from longtime teaching expert Maha Bali, a professor of practice at the Center for Learning and Teaching at the American University in Cairo.
The Opposite of Cheating Podcast
https://open.spotify.com/show/5fhrnwUIWgFqZYBJWGIYml
(Produced by the authors of the book with the same name) the podcast shares the real life experiences, thoughts, and talents of educators and professionals who are working to teach for integrity in the age of AI. The series features engaging conversations with brilliant innovators, teachers, leaders, and practitioners who are both resisting and integrating GenAI into their lives. The central value undergirding everything is, of course, integrity!
Teaching in Higher Ed podcast, “Cultivating Critical AI Literacies with Maha Bali”.
https://teachinginhighered.com/podcast/cultivating-critical-ai-literacies/
In the episode, host Bonni Stachowiak and guest Maha Bali, a Professor of Practice at the American University in Cairo, explore the complexities of integrating artificial intelligence into higher education.
Bali advocates for a critical pedagogical approach, rooted in the work of Paulo Freire, urging educators to actively experiment with AI to understand its limitations and biases. The discussion highlights significant issues of cultural and implicit bias within AI systems. Bali provides concrete examples, such as AI generating historically inaccurate information about Egyptian culture, misrepresenting cultural symbols, and defaulting to stereotypes when prompted for examples of terrorism.
The Actual Intelligence podcast
speakswith Dr. Robert Neibuhr from ASU regarding his recent article in Insider Higher Ed: “A.I and Higher Ed: An Impending Collapse.” Full Podcast: https://podcasts.apple.com/us/podcast/is-higher-ed-to-collapse-from-a-i/id1274615583?i=1000725770519
with Bill Gates having just said that A.I. will replace most teachers within ten years, it seems essential that professional educators attune to the growing presence of A.I. in education, particularly its negative gravitational forces.
“A Map Makes You Smarter. GPS Does Not.”: A Story About AI, Work, and What Comes Next with Jose Antonio Bowen
Jose Antonio Bowen is introduced as a Renaissance thinker with a jazz soul. His background includes leadership roles at Stanford, Georgetown, and SMU, as well as being the president of Johnstreet College. He is also a jazz musician who has played with legends, a composer with a Pulitzer-nominated symphony, and the author of “Teaching Naked,” 30% off with the code TNT30 at Wiley “Teaching Change,” and “Teaching with AI.” 30% off Teaching Change or Teaching with AI with Code HTWN at JH.
He provided a workshop for us on AI Assignment and Assessments, where he mentioned:
“A map makes you smarter. GPS does not.”
It was such a small, quiet moment, but it cracked open something bigger. Because this wasn’t just about directions. It was about how we’re all starting to think less, remember less, and—if we’re not careful—become less, all thanks to the technology we depend on.
The Decline of Entry-Level Everything
Dr. Bowen shared that Shell, a global energy giant, had laid off nearly 38% of a particular workforce group. Internships? Vanishing. Entry-level jobs? Replaced.
Replaced by what?
Artificial Intelligence
Tasks that used to belong to interns or fresh graduates—writing reports, creating slide decks, analyzing data—are now handled by machines that don’t take lunch breaks or need supervision.
And that’s where the real twist came in: the people who still have jobs? They’re not the ones who can do the task better than AI. They’re the ones who can think better than AI. Who can improve, refine, and oversee what AI produces.
If AI is writing the first draft, the humans left in the room better know how to write the final one—with nuance, clarity, and insight.
Offloading Our Minds, One Task at a Time
Back to that GPS quote. Dr. Bowen called it “cognitive offloading”—how we gradually stop using certain mental muscles because tech is doing the lifting.
We used to memorize phone numbers, navigate with paper maps, even mentally calculate tips at restaurants. Now? We ask Siri.
The scary part isn’t that we’re forgetting how to do these things. It’s what happens when we offload creativity, problem-solving, and thinking itself.
Because if AI can be creative—can write poems, code apps, design marketing plans—what do we do? What’s left for us?
Creativity, Reimagined
But here’s where things got interesting. Dr. Bowen isn’t anti-AI. In fact, he practically gushed about it.
He showed how AI can be used to spark creativity, not stifle it.
He explained how students could upload a 700-page textbook and have the AI turn it into a podcast. A nine-minute podcast. With baseball analogies, if that’s what helps them learn.
He talked about using AI to create personalized assignments: instead of a generic math problem about trains, give a politics student a question about voter turnout rates. Suddenly, they care. Suddenly, they’re engaged.
Because AI isn’t replacing the teacher—it’s becoming the chalk, the blackboard, the entire toolset that a smart educator can use to make learning come alive.
Prompt Like a Pro
Here’s another nugget that stuck with me: prompting isn’t coding. It’s storytelling.
Don’t just ask the AI to “fix your proposal.” Ask it to “transform your proposal into something your provost will love.”
Use emotion. Use intent. Give context. AI, it turns out, responds best when it knows what you’re really trying to say.
The 70% Problem
Still, AI isn’t perfect. Dr. Bowen introduced what he called the “70% problem.”
AI can do a lot of things—but only up to a C-level standard. That’s fine for a rough draft. It’s dangerous for a final product.
If students rely on AI to do the work, and they can’t take it past that 70% mark, then what happens when employers expect more?
The solution? Raise the bar.
What used to be acceptable for a B or C should now earn an F—unless the student can make the AI’s work better, smarter, more human.
From Tools to Teaching Assistants
The future of education, the he argued, is not about banning AI—it’s about designing with it.
He showed how teaching assistants could use AI notebooks filled with chemistry texts to answer student questions on the fly.
How AI can test business plans, simulate presidential decisions, or offer critiques from the perspective of a political opponent.
How students can train AI to “be” Einstein and ask it about thermodynamics at their own pace, in their own language.
AI isn’t replacing teachers—it’s becoming part of the classroom, like textbooks once were.
The Arms Race
Of course, there’s a darker side. AI can cheat. It can take online courses for students, fake typing patterns, even simulate human error.
Dr. Bowen called it an “arms race” between those building smarter AI and those trying to prevent it from being misused.
But even in this, he saw hope.
If educators embrace AI—not as an enemy but as a creative partner—they can design assignments AI can’t complete alone. They can build simulations, storytelling challenges, and editing tasks that require a human mind.
Because at the end of the day, that’s what this moment demands: humans who think more deeply, ask better questions, and create things worth remembering.
Final Words
The session ended with a simple truth:
“AI raises the floor. You must raise the ceiling.”
Whether you’re a student, a teacher, a manager, or a job-seeker, AI is now the baseline.
It will write the first draft, sketch the first idea, solve the first problem.
But it’s still up to us to bring the brilliance.
AI can produce work at a “C” level, which is problematic if students can only perform at that level. Instructors need to raise their standards and expectations. Assignments that would have been considered a “C” should now be evaluated as an “F” if they only meet the level of quality that AI can produce.
Implications
Students need to surpass AI capabilities to be competitive in the job market, especially in fields like coding and writing.
And maybe—just maybe—it’s time we all learned to read the map again.