The short answer
To choose an AI literacy curriculum for middle school, look past which chatbot it uses and judge the thinking it builds: whether students learn to question outputs, understand how data and models work, spot error and bias, protect privacy, and keep humans responsible for decisions. Score candidates on those criteria, check whether it requires student AI accounts or continuous screens, and confirm it connects lessons into a sequence rather than a one-off activity.
Score any AI literacy curriculum on six criteria
1. Judgment over tool use
Does it teach students to evaluate and question AI outputs, or mostly how to prompt a tool? Judgment is the durable skill; prompting techniques change with every model.
2. How data and models actually work
Does it make data, patterns, training, and limitations concrete — ideally with physical or visual models — rather than treating AI as magic?
3. Error, bias, and verification habits
Does it build a repeatable habit of checking outputs for error, bias, and missing context, and asking what evidence would confirm a claim?
4. Privacy and account requirements
Does the core curriculum require individual student AI accounts and continuous screens, or can it be taught with adult-controlled or offline materials that protect student data?
5. Human responsibility
Does it teach who stays accountable for decisions — where a human must review, escalate, or override an AI system rather than defer to it?
6. Sequence and depth
Is it a connected sequence where later lessons build on earlier evidence, or a one-off lesson? Depth and transfer come from sequence, not from a single exciting activity.
How to run the evaluation
- Rate each of the six criteria 0 (absent), 1 (partial), or 2 (strong), for a 12-point total.
- Treat judgment (criterion 1) and verification (criterion 3) as must-haves; a curriculum weak on both is not really teaching AI literacy.
- Ask to see one real lesson. A strong sample should make students identify the task, inspect an output, name a possible failure, find evidence, and decide what a person must review.
- Run a ten-minute verification drill: give the learner one confident AI-style claim, two possible sources, and one missing-context clue; then observe whether the curriculum's routine helps them verify rather than guess.
- Check the age fit, weekly time, adult role, privacy rules, account requirements, and the student evidence produced—not just the topic list.
- A 9-or-higher score is a useful shortlist signal, not proof of quality. Read the lesson sequence and confirm that later work actually reuses earlier evidence.
How EdReal's AI Literacy Lab scores on these criteria
EdReal's AI Literacy Lab is a 12-week program for ages 8–18 built around exactly these criteria: students inspect how AI uses data and patterns with printed system maps and cards, test outputs for error and bias, practice evidence checks, and design human-in-the-loop and privacy rules. Live student AI accounts are not required, and the sequence builds toward a responsible-use capstone. It is a supplementary program, not a replacement for core coursework.
See Your Child uses eight everyday parent observations to highlight durable skills, offer a free 15-minute activity, and recommend a Lab only when the fit is clear.
Important limits
This rubric is a planning aid, not an accreditation standard, and different settings weight the criteria differently — a school may prioritize privacy and standards documentation while a family may prioritize screen posture. No single curriculum is best for everyone.
EdReal Labs are supplementary, inquiry-based learning experiences designed to complement core academic work. They do not claim accreditation, formal district adoption, or replacement of core coursework.
Frequently asked questions
Is prompting the most important AI literacy skill?
No. Prompting is one practice inside a larger set that includes questioning outputs, understanding data and models, checking for error and bias, protecting privacy, and keeping humans responsible. A curriculum built only around prompting ages quickly as tools change.
Does good AI literacy require students to use ChatGPT or other AI accounts?
No. Strong AI literacy can be taught with adult-controlled demonstrations or offline materials — printed system maps, example outputs, and scenarios — which also avoids putting student data into a live service.
What age is AI literacy appropriate for?
The core questions can be introduced in upper elementary and revisited with more depth through high school. EdReal's AI Literacy Lab spans ages 8–18 by adjusting reading level and the expected depth of evidence.