The AI Usage Scale for Assignments and Syllabi. One Clear Framework for Every Assignment
"My students use AI for everything."
"How do I AI-proof my assignments?"
"How do I stop AI from doing the homework for them?"
I hear some version of these three stories every time I sit down with teachers. The assignment took an evening to plan, and it was completed after a ten-second conversation with a chatbot.
How do you handle this as a teacher? Does your school have a clear AI policy in place that supports your decisions? Have any expectations been set in your classroom that provide students with clear guidance when it comes to AI usage expectations?
These conversations can't wait any longer. The AI conversation has moved from theory to practical classroom application. Unfortunately, most schools and teachers I work with aren't prepared for actual classroom implementation.
So, in my typical fashion, I made something for it: a four-level AI usage scale that puts one clear AI label on every assignment and every syllabus, setting the stage for teachers to provide clear guidance and classroom expectations. It's free, and the full breakdown is below.
Your students are already using it.
The numbers match what teachers are telling me. In a fall 2025 survey of U.S. teens ages 13 to 17, 54% said they had used an AI chatbot to get help with schoolwork, and about one in ten said they do all or most of their schoolwork with a chatbot's help (McClain, Anderson, Sidoti, & Bishop, 2026). In the same survey, 59% said students at their school use AI to cheat at least somewhat often.
Here's the part that matters for a teacher: most of those students are taking AI usage into their own hands because expectations aren't clear. One teacher bans AI, the next one encourages it, and the assignment instructions rarely guide students on appropriate use. When the expectation lives only in the teacher's head, students fill the gap with their own judgment.
A label only works when students know what it means. That is why the scale starts with a conversation, long before the first assignment gets a level. Talk with your students about what AI does well, where it gets things wrong, and why some work has to be theirs alone. Then keep that conversation going all year: when a new tool shows up, when a student asks whether something is allowed, and every time you hand out an assignment with a different level on it. By the time a student sees "Level 2" at the top of a page, they already know what you expect and why.
This is the part I walk teachers through in my workshops, because it changes what AI does in a classroom. When the expectations are clear, AI can sharpen a student's critical thinking and pull them deeper into the work, and students learn to keep the hard thinking for themselves.
One label, written on the assignment, answers the question before a student has to guess.
The Olive & Rose Education scale is my K-12 version: four levels, plain language, and a short statement students attach to their work so they tell you exactly what they did.
The four levels
Level 1: AI-Free
Student role: Work independently. No AI use allowed.
Teacher action: Prohibit all AI tools. Require an academic honesty pledge if needed.
Use it for: Assessing pure student understanding and skill.
When and why: Use AI-Free when you introduce new content or material. Students need to learn to walk before they run, so they show you they have mastered a skill on their own before AI is allowed to help with it. It is also the right level for any assessment where you need to see exactly what a student can do.
Example: After the first lessons in an argumentative writing unit, students write one paragraph in class, on paper, using two pieces of evidence from an article the class read together.
Honesty pledge example: "I affirm that the work I am submitting was completed entirely by me without any assistance from artificial intelligence tools. I understand that using AI in any form on this assignment would violate the academic integrity expectations of this course."
Level 2: AI-Assisted
Student role: Use AI for planning, brainstorming, or feedback only. Final work must be original.
Teacher action: Clarify which AI tasks are permitted. Require a short AI usage statement.
Use it for: Supporting thinking without letting AI generate content.
When and why: Use AI-Assisted once students are practicing a skill and a thinking partner would help, while the finished work still needs to be entirely theirs. AI can question a student's idea, help organize a plan, or give feedback on a draft, and it writes nothing the student turns in. This part is crucial because AI tools will typically ask the user: "Would you like me to revise the draft to incorporate these ideas for you?". This is where teachers can step in and guide students to not take the conversation that far and explain the why behind it.
Does that ensure students won't use AI to draft the final deliverable for them? Not at all. But a conversation on expectations and ethical AI use is much better than a student being penalized because no expectations were set.
Example: A student starts with a broad research topic, goes back and forth with AI at least three times to narrow it, then writes the final research question in their own words.
Usage statement: "For this assignment, I used AI tools only for the following approved purposes (e.g., brainstorming, planning, or receiving feedback). No AI-generated content was included in the final submission."
Level 3: AI-Create
Student role: Collaborate with AI to analyze or create content.
Teacher action: Teach critical use of AI. Require disclosure and human oversight of the content.
Use it for: Co-creating with AI while students remain the primary authors, such as generating group discussion questions or images.
When and why: Use AI-Create once students have demonstrated a baseline understanding of the content. At this level, AI's words can end up in the final work, so the student's job becomes directing revisions, checking facts, and owning the result. You can see their thinking in the choices they make with AI's draft.
Example: Students ask AI for a first draft of a study guide, direct at least two rounds of revision, then fact-check every claim before sharing it with the class.
Usage statement: "For this assignment, I used AI tools only for the following approved purposes (e.g., analyzing output, co-creating content). Some AI-generated content was included in the final submission."
Level 4: AI-Empowered
Student role: Fully integrate AI to create advanced, creative solutions.
Teacher action: Encourage innovation. Require accuracy checks and detailed AI usage documentation.
Use it for: An interview with a famous person, practice tests, or real-world simulations like a mock debate or a Socratic discussion.
When and why: Use AI-Empowered for learning experiences AI makes possible: simulations, interviews, debates, and practice tests. AI generates most of the content here, so checking it becomes the assignment. Students verify what AI said against real sources and document what they found.
Example: Students interview a historical figure through an AI tool, asking five to seven questions about the person's motives and decisions, then check one claim against a primary source and write whether it held up.
Usage statement: "For this assignment, I used AI tools only for the following approved purposes (e.g., creating dialogue with AI). Mostly AI-generated content was included in the final submission."
How to put it on an assignment this week.
Pick one assignment you're giving in the next two weeks.
Ask what you're actually assessing. If it's the student's own skill (a timed paragraph, a math procedure, a reading response), label it Level 1. If it's their thinking and AI can be a sparring partner, try Level 2.
Write the level at the top of the assignment sheet, right next to the grading criteria.
Paste the matching pledge or usage statement at the bottom and make it part of the submission.
Spend two minutes on it when you hand the assignment out. Explain to your students why the selected tier was given to the assignment and what AI usage is expected.
For the syllabus, one line does the job: "Every assignment in this course carries an AI level from 1 to 4. If an assignment has no level, ask me before you use AI."
What happens after the first assignment.
The scale sets the expectation. Levels 3 and 4 ask the teacher to teach critical use of AI and to require accuracy checks, and that takes practice with the tools your school has approved. The labels also mean the most when every teacher in the building uses the same four. A student who sees "Level 2" in English and a different system in science is back to guessing.
You're not behind if your school hasn't settled this yet. Most of the schools I talk with are right where you are.
Questions school leaders ask
Should we require the scale schoolwide or let teachers opt in?
Schoolwide works better. Students move between six or seven teachers a day, and a shared vocabulary is what makes the labels mean something. If your faculty needs a smaller start, begin with one department or one grade level and widen it after a quarter.
Does Level 1 mean we need AI detection software?
The scale doesn't depend on detection. Level 1 rests on the pledge, on doing the work in class when it matters most, and on the conversation you have when you hand the assignment out. AI detection software is not completely reliable, so we do not encourage its use. We encourage guiding students on appropriate AI usage that enhances a student's learning experience.
What grade levels is it for?
Any grade where students can reach generative AI tools on their own. Check the age requirements of the tools your school has approved; UNESCO recommends an age limit for independent conversations with generative AI platforms (Miao & Holmes, 2023).
Your next step
If you want your whole faculty using the scale the same way, the Educator AI Essentials Workshop is two hours with every teacher in the building, on your school's approved tools, and it includes an introduction to the scale. You can see it on our Faculty Professional Development page or book a free consultation and we'll talk through what your teachers need.
Which assignment are you going to label first?
References
McClain, C., Anderson, M., Sidoti, O., & Bishop, W. (2026). How teens use and view AI. Pew Research Center. https://www.pewresearch.org/internet/2026/02/24/how-teens-use-and-view-ai/
Miao, F., & Holmes, W. (2023). Guidance for generative AI in education and research. UNESCO. https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research