Student learning · Grades 9–12 · 50 minutes
The Offloading Audit
Students audit a week of their own AI use, sort each instance into thinking worth keeping and busywork worth delegating, and confront overreliance, writing personal rules for when to struggle and when to let the machine carry it.
The debate in the news swings between "AI will hollow out a generation's thinking" and "it's just the next calculator." This activity puts the question where students can actually answer it: in their own week. Students log every place they used (or wanted to use) AI for schoolwork, then audit each one, was the AI carrying thinking that was the whole point of the task, or handling genuine busywork around the real learning? They meet cognitive offloading as a concept with a real research tension (some studies tie heavy reliance to weaker recall and effort, while careful use as a tutor or checker can support learning) and sort their own log against it. The goal is not guilt or a ban; it is judgment. Students name where overreliance is quietly costing them the skills they came to build, and where delegation is smart, then write rules they can keep and test. It is the ELE question aimed at themselves: what evidence shows the tool improved my learning, not just my output?
In this packet
- Handout A: My AI Week (organizer)
- Handout B: The Offload Line (worksheet)
- Handout C: What the Evidence Says (reading)
TCEA ELE indicators
- AI4.1 I know how to promote critical thinking skills in relation to AI-generated information.
- S4.1 I know how to take responsibility for my own learning.
- S4.3 I am aware of my learning strengths and areas for improvement.
- AI4.3 I am aware of strategies to develop AI-related problem-solving skills.
Full facilitator guide: mglearn.github.io/eles/activities/the-offloading-audit.html
Handout A for The Offloading Audit
My AI Week
For one normal week, log every time you used AI (or wanted to) for schoolwork. Be honest, this is private and not graded. Fill the last two columns later, in class.
| Class / task | What I asked AI to do | Did I understand it after? | Offload line verdict (fill in class) |
|---|---|---|---|
Handout B for The Offloading Audit
The Offload Line
Use your week's log to analyze your AI use, then write rules you can actually keep and check.
Roughly how much of your week's AI use CARRIED THE THINKING that was the point of the task, versus handled BUSYWORK around the learning? (You can draw two bars.)
Your clearest SMART delegation (AI handled busywork, your learning was fine):
Your riskiest OVERRELIANCE (AI carried thinking in a class built to grow that thinking). What skill is it costing you?
RULE 1 (name the class/situation, what you'll do, and how you'll check it worked):
RULE 2 (same: situation, action, evidence you'll look for):
RULE 3, optional:
Handout C for The Offloading Audit
What the Evidence Says
Read it. Underline the one finding that most matches your own experience. Mark anything stated as a LINK (correlation) versus a proven CAUSE.
1Researchers use the term cognitive offloading for handing a mental task to an external tool, a calculator, a map app, a search engine, and now AI. Offloading is ancient and often smart: nobody memorizes every phone number anymore. The question is always what you are offloading and whether you needed to learn it.
2Some concerning findings: studies have linked heavy reliance on search and GPS to weaker memory for the information itself and poorer independent navigation. Early research on AI assistants points the same way, when people let AI generate answers for tasks they were meant to learn, they often recall less and report putting in less mental effort. These are associations, and most are short-term; they do not prove AI permanently damages thinking.
3Some reassuring findings: AI used well can support learning. As a tutor that asks you questions, a checker that reviews reasoning you did yourself, or a tool that handles busywork so you can focus on the hard part, it can deepen rather than replace thinking. The benefit shows up when the human keeps doing the thinking that is the actual goal.
4The pattern across both: the tool is not the whole story, the task is. Offloading arithmetic you already understand frees you to think about the problem. Offloading the reasoning you came to class to learn trades away the skill itself. The same app can do either, depending on how you use it.
5A fair bottom line: AI is neither a brain-rotting machine nor a harmless calculator. It is a powerful offloading tool, and whether it helps or costs you depends on a judgment you have to make, task by task. That judgment is what this audit is building.