Nine units, one arc. Each grade is a self-contained five-lesson unit that stands on its own — and together they carry a student from “is a machine alive?” at five years old to “when should I decline to use AI?” at fourteen.
Grade K — Smart Machines Are Still Machines
Some machines can seem smart, but people make them — they aren’t alive and don’t have feelings. Kindergartners sort the world into alive and not-alive, discover that a talking speaker belongs with the machines, and meet a class puppet who helps them practice one big truth: a talking machine isn’t a friend with feelings. The project: A class mural of machines and the people who make them. Safety thread: “A talking machine isn’t a friend — tell a grown-up.” Students name the trusted grown-ups they go to with feelings and worries.
Grade 1 — People Teach Machines
AI isn’t magic — it’s taught by people, with examples. And if the examples are bad, it can be wrong or unfair. First graders “train” a partner to sort shapes using example cards, watch the behavior change when the examples change, catch a “machine” calling a muffin a dog, and see how leaving a whole group out of the examples makes things unfair. The project: In groups, students build an example set that teaches a simple idea — happy/sad, day/night — and present how it teaches. Safety thread: People choose the examples, so people are responsible. Fairness means including everyone.
Grade 2 — AI Senses the World in Its Own Way
Machines sense with sensors — a camera for “seeing,” a microphone for “hearing” — and sensing is not the same as understanding. Second graders match human senses to machine sensors, discover what machines can’t sense at all, and learn that if a device can sense, it can record — which opens the first honest conversation about privacy. The project: A “Sensor Detectives” poster mapping the sensors hiding in everyday devices. Safety thread: “Some information is private — ask a grown-up first.” Students sort what’s fine to share with a device from what isn’t.
Grade 3 — AI Learns From Patterns
AI learns from patterns in lots of examples — and data comes from somewhere. Third graders become pattern-finders, training simple classifiers by hand and tracing where training data actually comes from — real people, real choices. The project: A hand-built pattern machine students train, test, and explain. Safety thread: Data comes from people — including you. Start noticing what gets collected.
Grade 4 — Good Data In, Good Results Out
Biased data isn’t just a technical problem — it’s unfair to real people. Fourth graders test what happens when training data leaves people out or tilts one way, then practice spotting bias and fixing it with better, fairer examples. The project: A fairness audit: find the bias, show who it hurts, and repair the data. Safety thread: Unfair data hurts real people — and people can fix it.
Grade 5 — Guesses, Not Facts
AI makes predictions — guesses, not facts. Verify what matters, and protect your privacy. Fifth graders see prediction from the inside: why confident answers can still be wrong, how to check claims against real sources, and what personal information deserves protection online. The project: A verification challenge — students fact-check AI-style claims and show their evidence. Safety thread: Verify before you trust. Guard your personal information.
Grade 6 — How Generative AI Makes Things
Generative AI creates text and images — and sometimes makes things up. Hallucination, bias, and deepfakes are real. Sixth graders look under the hood of generative AI using curated, printed examples: where invented “facts” come from, how bias shows up in generated content, and how to spot manipulated media. The project: A spot-the-fake portfolio: students analyze real and generated samples and explain the tells. Safety thread: If it can be generated, it can be faked — check before you share.
Grade 7 — It Talks Like a Person. It Isn’t One.
AI that talks like a person isn’t a person — this is the program’s dedicated AI-companion safety unit. Seventh graders examine how companion chatbots are designed to feel like friends, why that design deserves healthy skepticism, and how to set boundaries — including the most important one: real problems go to trusted adults, not to software. The project: Students build a boundary guide — realistic scenarios, healthy responses, and where real help lives. Safety thread: The lead safety grade of the program. It isn’t a person; real problems go to trusted adults.
Grade 8 — Responsible Use — the Capstone
Using AI responsibly, honestly — and deciding when not to use it at all. Eighth graders finish the arc as capable, honest users: fact-checking AI output against sources, disclosing AI use with integrity, and — the capstone judgment — recognizing the work that should stay fully their own. The project: The capstone: a real task completed with documented, honest, defensible AI decisions. Safety thread: Honesty about AI use — and the judgment to sometimes decline it.