Designing Lessons That Use AI Without Leaning on It
One of the trickiest parts of bringing AI tools into a classroom is making sure students still do the thinking. A few approaches that seem to work well in practice:
- Have students critique an AI-generated answer rather than simply accepting it.
- Use AI tools for brainstorming, then require original synthesis.
- Ask students to compare their own reasoning to the tool’s output.
The goal isn’t to avoid AI tools, but to design assignments where the tool is a starting point rather than a shortcut.
Sequencing matters more than the tool itself
Most of the friction around AI in classrooms isn’t really about the technology — it’s about when in the learning process a student reaches for it. A student who struggles through a first attempt, hits a wall, and only then consults an AI tool to see another approach is doing something fundamentally different from a student who opens the tool first and transcribes the output. Same tool, same assignment, completely different cognitive experience. Teachers who’ve had success here tend to build that sequencing directly into the assignment structure, rather than leaving it to student discretion. A “your attempt first, then the comparison” requirement does more work than any honor-code-style policy about AI use.
A few approaches that seem to hold up well in practice:
- Have students critique an AI-generated answer rather than simply accepting it — asking them to find where it’s wrong, imprecise, or missing context.
- Use AI tools for brainstorming, then require original synthesis — the tool expands the option space; the student still has to choose and build.
- Ask students to compare their own reasoning to the tool’s output — not just whether the answers match, but where and why they diverge.
Grading the process, not just the artifact
Assignments that ask students to show where their reasoning diverged from an AI’s — and to explain why one approach was better, worse, or just different — shift the grading target from “did you produce a correct answer” to “can you evaluate reasoning, including your own.” That second skill is harder to fake, and harder to shortcut with a tool.
A caution worth naming
None of this works if the critique step becomes its own rote exercise. Students quickly learn to perform the appearance of critical engagement — a generic line about the AI “missing nuance” — without doing much real evaluation. The strongest versions of these assignments ask for something specific and checkable: which step was wrong, what assumption caused it, what a better version would look like, rather than an open-ended reflection paragraph.
Why this matters beyond the assignment
The goal isn’t to avoid AI tools, but to design assignments where the tool is a starting point rather than a shortcut. Used this way, AI functions less like an answer key and more like a second opinion a student has to argue with — a genuinely useful habit of mind, and one that’s arguably more transferable than the subject-matter content of any single lesson.
For educators evaluating how to bring AI into a course, the useful question isn’t “should students be allowed to use it,” but “at what point in the thinking process does this tool enter and does the assignment still require the student to do the part that mattered.”
