The Friday Download: AI Broke the Pop Quiz (And Might Save Assessment) (April 10, 2026)

The Friday Download — Show Notes "The Robot Wrote My Essay (Or Did It?)"

This week on The Friday Download, JR asks the question that's haunting every teacher, professor, and parent in 2026: did my student write this — or did their robot?

In The Big Weird, we dig into what the data actually shows about student AI use. Spoiler: over 90% of college students are using AI somewhere in their workflow, but the "everyone is cheating" story turns out to be way more complicated. We also talk about why AI detectors failed spectacularly — flagging human writing, missing obvious bot output, and disproportionately targeting non-native English speakers — and why institutions are backing away from them fast.

In Wait… That's Actually Cool, we explore the educators who are responding not by chasing cheaters, but by redesigning the assignments themselves. AI-vulnerable tasks (generic essays, cookie-cutter prompts) versus AI-resistant tasks (oral checkpoints, portfolio-based work, assignments tied to lived experience) — and why trying to build the second kind is accidentally producing better education than we had before.

And in The Tiny Tech Snack, five terms you need to know right now: AI-resistant assessment, process-based grading, oral checkpoints, AI disclosure, and why AI-proof doesn't mean tech-free.

Whether you're a teacher redesigning your syllabus, a student figuring out where the line actually is, or a parent wondering what your kid's school is doing about all this — this episode is for you.

🎙️ Hosted by JR DeLaney, The AI Learning Guide

REFERENCES

1. Lee, S. (2026, February 12). Has AI made academic cheating worse? 2026 data. PlagiarismCheck.org. https://plagiarismcheck.org/blog/has-ai-made-academic-cheating-worse-2026-data/
2. College Board. (2026, February 24). Faculty express near-universal concern that student AI use undermines academic integrity [Press release]. https://newsroom.collegeboard.org/new-college-board-research-faculty-express-near-universal-concern-student-ai-use-undermines
3. Roschelle, J. (2026, March 8). Real-time data shows exactly how students use AI on school technology. Education Week. https://www.edweek.org/technology/real-time-data-shows-exactly-how-students-use-ai-on-school-technology/2026/03
4. OpenEduCat. (2026, March 14). AI and academic integrity: A practical guide. OpenEduCat. https://openeducat.org/articles/ai-academic-integrity-guide-for-schools/
5. Northern Michigan University Center for Teaching and Learning. (n.d.). Creating AI-resistant assignments, activities, and assessments: Designing out academic dishonesty. NMU. https://nmu.edu/ctl/creating-ai-resistant-assignments-activities-and-assessments-designing-out
6. Dellarocas, C. (2026, February 18). AI will break assessment before it fixes it. The Credential Crisis. https://futurecredentials.substack.com/p/ai-will-break-assessment-before-it

Transcript

SPEAKER_00 0:03

My room was smart, but it's the moment socks. Now my toast is trying to pick the locks. It's the Friday download. This has been a week where teachers, professors, and parents all ask the same question. Did you write this? Or did your robot? Hey friends, I'm your AI learning tour guide, JR, and today we're diving into how AI is breaking our old school assessments. And weirdly, might be the thing that finally forces us to build better ones. We're talking cheating, chatbots, and why the pop quiz as we know it may not survive 2026. Alright, this is Mike. It's let's go! Let's start out with the cheating apocalypse that wasn't. Exactly. So let's rewind to the last couple of years. Cheating headlines were everywhere. If you only read the clickbait, you'd think every student on Earth was feeding entire assessments into a chatbot, pressing enter, and then wandering off to binge Netflix while the robot earned them a degree. And honestly, some did. But when people actually started looking at the data, the picture got a lot messier. Surveys now suggest that the vast majority of college students, over 90% in some samples, are using AI somewhere in their study workflow, but only a much smaller slice admit to actually using it to fully complete an assignment for them. So if you're picturing a world where every essay is 100% bot written, that's not really what's happening. Most students are doing something more human. They're using AI like we use Google, SparkNotes, or that one really smart kid in the group chat. Explain this concept. Give me practice questions. Help me brainstorm ideas. It's still messy, and yeah, some of that crosses into cheating. But it's not one big monolithic Everyone is lying story. Here's where it gets even weirder. When researchers ask students, hey, do you think turning an AI written work as your own is cheating? A big chunk says, Oh yeah, absolutely, that's that's cheating. And then a not tiny percentage said, Yeah, I know, and I've done it. Even better. A large majority say they're pretty sure they won't get caught. That's the teenage brain in a nutshell. This, I know this is wrong, but also I am invincible. Meanwhile, schools try to respond with AI detectors. For a while we saw this arms race. On one side we had students with jetbots, on the other side, instructors with is this AI detectors? That felt a little like horoscope apps for essays. They would flag perfectly human writing as well, probably AI. They would miss obviously AI-generated content. They also tended to be biased toward calling non-native English speakers fake. So, institutions started backing away. You'll see more and more policies saying we might use these tools as one piece of evidence, but we cannot rely on them as proof. Which leads to this very 2026 scenario, where the student says, I swear I wrote this. Teacher, the AI detector says you didn't. Student says, So a robot is accusing me of using another robot. Hmm. At some universities, both the suspected cheating and the accusation were mediated by AI. The essay came from a chatbot, the evidence came from another chatbot. And at some point, the humans had to step back in and say, okay, this is ridiculous. We need a different approach. And one more twist to this big weird segment. When you look at early cheating data, you see higher AI misuse rates popping up in under-resourced schools or high-pressure environments. If your choices are pay for a private tutor your family can't afford, or open a free chatbot that will happily explain every physics problem at 2 a.m., that's not hard to see how the temptation grows. So, yes, individual choices matter, ethics matter, but we're also looking at a system problem. Huge pressure, high stakes, and an AI tool that is always awake and never charges by the hour. But put that all together and you'll see why this cheating has become the most exhausting question in education. Which brings us to a better one. Maybe instead of asking, how do we watch and catch cheaters? we should be asking, why are our assessments so easy for a robot to fake in the first place? Well, let's move on to segment two with wait, that's actually cool. How AI is forcing better and more human assessments. So here's the twist I love. A lot of people in the higher ed are now arguing that the biggest risk of AI isn't just cheating, it's the possibility that we keep pretending our old assignments still work. Think about it. If an AI can write your standard five-paragraph essay better than your students can, then that well might say something about the assignment as much as it does about the AI. We built decades of schooling on tasks that were easy to grade, easy to copy, and now very easy to automate. Like summarize this chapter, explain this theory, do 20 nearly identical math problems. Those were never great measures of deep learning. They were measures of can you follow the formula? Oh, and AI is excellent at formulas. So some educators are doing something really interesting. They're starting to classify tasks as whether AI vulnerable and or AI resistant. AI vulnerable tasks are the ones the chatbot can nail in seconds. Generic summaries. Basic definitions. Cookie-cutter essays on overuse topics like should school uniforms be allowed? AI-resistant tasks shall still allow AI in the mix, but they require human context, judgment, or performance that the tool can't fake. Things like tying your content to your own lived experience, working with local data or real-world constraints, presenting and defending your thinking live, showing your process over time instead of just submitting a polished final product. Let's make that concrete. Old history prompt says explain the causes of World War One. AI can do that in three seconds. And probably throwing a quote here or there for some style points. A more AI-resistant prompt might better say, use at least one story from your own family or community through letters, photographs, interviews, and connect it to one of the major causes or impacts of World War One. Include where you got your information and be ready to share and answer questions in class. Now, AI can still help. It can explain the causes, it'll help you refine your writing, but it cannot invent your grandmother's letters from 1917. It cannot stand in front of your class yet and answer follow-up questions about your granddad's story. Or think about STEM. Here's the old assessment with STEM. Here's a problem set solved for X 20 times. And I eat that for breakfast. Newer style assessments might look like design a simple prototype to solve a real problem in your school. Traffic in the hallway, too much noise in the cafeteria, or really anything. Document your design process, including failures, and present your trade-offs to the class. Again, AI can brainstorm ideas or check your math, but it can't build your prototype for you yet. Or show what might actually happen when you tested in your hallway, or answer live questions from your classmates. We're also seeing more oral exams and many conferences, recorded video reflections, drafts and journals graded as part of the assignment, portfolios that accumulate over time. All of these things do two things at once. First, they make cheating with AI harder, and then they often give a better picture of what the student can actually do. Now, on the policy side, there's a quiet but important shift as well. A lot of schools and universities are moving from ignore AI or ban AI to let's define what counts as acceptable AI support, what requires disclosure, and what clearly crosses the line. Some examples of where this is landing. You can use AI to brainstorm ideas or get explanations as long as the final work is yours. You must disclose significant AI help, like if you asked it to rewrite a full paragraph. You may not submit AI-generated work as if you wrote it, and you may not lie about how you use a tool. And crucially, they're writing these expectations down in plain language so students aren't guessing what's isn't is not allowed. You're also seeing more institutions say, we won't accuse someone of cheating based on what an AI detector says. We need real evidence. That kind of clarity is good for everyone. It reduces the paranoia, it invites transparent use of AI as a tool, and it puts the focus back on what actually matters. The can you think? Can you learn? Can you do something meaningful with the knowledge, not just generate words about it? And the coolest part for me is in trying to design AI-resistant assessments, a lot of educators are accidentally designing better assessments, more authentic, more connected to real life, more focused on higher order thinking and creativity. The robot forced us to admit we can do better than the worksheet. Now let's move on to our final segment of Tiny Tech Snack Bites. Alright, so we have three bite explainers to make you sound smarter in your next staff meeting, parent teacher conference, or group chat meltdown. And if you have water coolers still at work, we'll fill that in too. First, AI resistant assessments. We talked about this. This is an assignment designed so that AI can support you, but it cannot do the whole thing for you. Think of it like baking a cake. AI can be your recipe, your sous chef, your reminder not to burn it. But at some point, someone has to actually crack the eggs and put the thing in the oven. So why does this matter? Well, when tasks require your personal experience, real-world context, or in the moment explanation, grades start to reflect your learning again, not just the quality of your prompt. Our second one, process-based grading. Instead of only grading the shiny final essay or project, your teacher also grades your proposal, your outline, your drafts, your reflections along the way. We're not just asking, where did you land? We're asking, how did you get there? So why? Well, it is much harder to outsource a whole evolving process to AI. And from a learning perspective, this is where the growth lives, right in the messy middle where you're still trying to figure things out. And snack three, oral checkpoints. These are short, low pressure conversations or mini presentations where you explain your work to a human, your teacher, a panel, even your fellow classmates. Think of it like talk with me how you solve this problem, instead of just hand it in and walk away. Why does this matter? Well, a student who actually understands what they turned in can talk about it, tweak it, defend it. A student who copy paste it from a chatbot hits a wall about three follow-up questions down the line. Yeah. Let's go with AI disclosure. Instead of never used AI, or use AI, but don't you dare tell me. This is a simple classroom rule. Like you are allowed to use AI for brainstorming and revision. If you do, you must say where and how you used it. Why? Well, disclosure turns AI from a secret weapon into a visible tool. That's exactly what students will need in real jobs, using powerful tools openly and ethically, not sneaking them under the table. And our final snack, AI proof doesn't mean tech proof. Designing AI resistant assignments is not about pretending AI doesn't exist or banning everything with a login screen. It's about asking how can technology help students think deeper, not think less? Why? Well, if we get this right, AI becomes the pressure that pushes us to drop the worst busy work and double down on the work that actually matters in life, those projects, collaborations, creativity, and critical thinking. So if you've ever stared at a suspiciously polished essay and thought, there is no way a sleepy teenager wrote this at 11 59 p.m., you are very much not alone. The old question, is this cheating, is still important. But the bigger, better question is, is this assessment worthy of a world where everyone has access to AI? For teachers out there, you don't have to outsmart the bots. You just have to ask better questions. Questions that require real thinking, real context, and a real human voice put to it. For all you students out there, AI can absolutely be your study buddy. But if it's doing all the work, it's also stealing all your learning. And future you, the one sitting in a job interview or at a lab or a boardroom, is going to notice. And for your parents and leaders out there, don't just ask, is my kid allowed to use AI? Instead, ask, how is their school designing assessments so that with or without AI, my kid is actually learning something real? And I'd love to hear from everyone. If you've got a wild, was this written by my student or by a robot story? Or if you've tried an assignment that worked better in the age of AI, pass it along. I would love to read it. We might even feature it on a future episode. Alright, that's our whirlwind tour of new assessment norms in the age of AI. This has been your AI Learning Guide, JR, and this has been your whirlwind replay of how AI broke the pop quiz and just might save assessments from itself. Until next time, may your prompts be honest, your rubrics clear, and may at least one essay this week sound convincingly like an actual teenager. Thanks for tuning in. Well, until the next one.