Teaching AI literacy — age-appropriate ideas for K–12

Students are using AI whether or not we teach them about it. RAND survey data found that the share of middle and high school students using AI for homework climbed sharply during 2025, and many students say they worry it’s affecting their thinking. AI literacy — understanding what these tools are, what they get wrong, and when they help — is quickly becoming a basic skill, and many states now encourage or require schools to teach it.

You don’t need to be a technology expert to start. Here are ideas by grade band. Check your school’s policy and the age requirements of any tool before students use it directly — many of these activities need no AI account at all.

Big ideas worth teaching at every age

1. AI learns patterns from lots of examples. It doesn’t “know” things the way people do.
2. AI can be wrong, even when it sounds sure. Checking matters.
3. AI reflects the data it learned from, including people’s biases.
4. People make choices about how AI is built and used — and students can too.
5. Some things should stay private. Don’t share personal information with chatbots.

Grades K–2: patterns and machines

• Sorting game. Give groups picture cards (animals, shapes) and have them invent a rule to sort them. Then show a card that “breaks” their rule. Talk about how computers learn rules from examples and can be confused by new things.
• Robot or person? Read short sentences and discuss: could a machine do this? (Count apples: yes. Know how your friend feels: not really.)
• Smart speakers. If students use voice assistants at home, talk about how they listen for a wake word and why we don’t tell them secrets.

Grades 3–5: predicting and checking

• The prediction game. Write the start of a sentence — “Once upon a…” — and have students predict the next word. Then try a less common start. Explain that AI chatbots work by predicting likely next words, very fast, with lots of practice. (Our explainer has the details for you.)
• Fact-check the robot. Show a short AI-written paragraph on a topic you’re studying that contains one or two errors (check it yourself first). Students hunt for the mistakes using their books.
• Who’s missing? Look at AI-generated images or examples together and ask who and what is shown — and who isn’t.

Grades 6–8: bias, sources and good use

• Same question, different answers. Show two AI responses to the same question. Which is better, and how can you tell? Introduce checking against reliable sources.
• Spot the made-up source. Show a list of citations, some real and some AI-invented. Students search for each one.
• Write the rules. Have the class draft guidelines for when AI use is helpful versus harmful for learning in your subject. This builds buy-in for your actual policy.

Grades 9–12: ethics, careers and critical use

• Policy debate. Should schools use AI to grade essays? Should AI-generated images be labeled? Students research and argue positions.
• Audit an output. Students use an approved tool to research a topic, then document every claim they verified, changed or rejected — and reflect on what they learned that the tool didn’t give them.
• Deepfakes and trust. Discuss how realistic fake images, audio and video affect news, elections and students’ own lives — including why sharing fakes of real people causes real harm.
• AI in careers. Investigate how AI is changing a field students care about, and which human skills matter more because of it.

Free curricula to borrow from

• Day of AI (MIT RAISE) — free K–12 AI literacy lessons: dayofai.org
• Code.org AI lessons and videos — student-friendly explanations of how AI works: code.org/ai
• TeachAI Toolkit — guidance for schools developing AI policies: teachai.org/toolkit

More vetted resources are on our resources page.

Sources

• eSchool News, Student use of AI for homework rises (reporting RAND American Youth Panel data, March 2026)
• Education Commission of the States, AI in education: state policy trends (May 2026)