What Is AI Literacy — and Why Does It Matter in Your Class?
AI literacy isn't about turning every teacher into a computer science instructor. It's about giving students the foundation to ask better questions, spot problems, and make thoughtful choices when AI is involved — which is already every day of their lives.
For Today
Students are already using AI tools, seeing AI-generated content on social media, and navigating AI recommendations in search results. Without guidance, they're learning from friends and trial and error.
For Tomorrow
AI will reshape every career path — not just tech jobs. Students who understand how to use AI critically and responsibly will be better positioned in any workforce they enter.
8 Things Every AI-Literate Educator Should Know
These aren't computer science concepts — they're the practical understandings that help you make good judgment calls about AI in your classroom and model those decisions for students.
AI models like the ones powering ChatGPT and Gemini were trained on roughly ten trillion words — books, websites, transcripts, legal documents, social media, and more. That's the equivalent of a human reading nonstop for 76,000 years. This scale is why AI can respond fluently on nearly any topic. It's also why students need to stay sharp: the more you know about a subject, the better you can judge whether AI actually got it right.
AI doesn't actually "know" things the way humans do. It makes statistical best guesses based on patterns in its training data — and it sounds confident even when it's wrong. These errors are called hallucinations. The practical implication: AI output always needs a human to verify it, and students who know more about a topic are better equipped to catch mistakes.
Developers tune AI models differently — adjusting tone, response length, how they handle sensitive topics, and more. This is why Claude, Gemini, and ChatGPT respond differently to the same prompt. A useful classroom activity: give students the same prompt to try in multiple tools and compare the results. This naturally raises AI literacy questions about where differences come from and which response is "better."
Many students don't distinguish between an AI assistant and a search engine — and increasingly, that distinction is harder to see as AI summaries appear at the top of Google results. Both can give wrong information; the mechanisms are just different. Teaching students to ask "Where is this coming from?" before trusting a result is an AI literacy skill that applies equally to search and AI chat.
Because AI is trained on human-generated data, it reflects human biases. This shows up in obvious ways (gender representation in AI-generated images of engineers) and subtle ones (AI tends to recommend dark coffee roasts, coastal vacation spots, and popular music genres over less common alternatives). Helping students notice and name these leanings in your subject area is a powerful critical thinking exercise.
Free AI tools often use conversation data to improve future models. Students should never enter personally identifiable information (PII) — names combined with other details, ID numbers, medical information, passwords — into AI prompts. A good rule of thumb: if you wouldn't want a hacker to find it, don't type it into AI.
When students go to AI first, they outsource the thinking — and that's precisely where the learning lives. Design activities where students brainstorm, draft, or reason on their own before bringing in AI. Then AI becomes a comparison tool or a way to push deeper, not a shortcut past the hard part. This maps directly to JA's "Who is doing the thinking?" framework.
AI can produce a polished essay faster than any student. If we only grade the final product, we've accidentally incentivized the shortcut. Build in opportunities for students to talk about their process — what they found hard, what they'd do differently, what surprised them. The struggle is where growth happens, and AI-era education needs to make that visible and valued.
"By the Way" Lessons
A "By the Way" lesson is a 30-second AI literacy moment woven into a conversation you're already having. No planning required. Just a nugget of truth about how AI works, connected to whatever your class is doing that day.
The Formula
Start with
"By the way, did you know…"
Add one truth
One short fact about how AI works
Connect it
Tie it to today's lesson or discussion
10 Ready-to-Use "By the Way" Lessons
Pick any of these when the moment naturally arises. Adapt the wording to fit your class.
AI makes mistakes. They're called hallucinations. Sometimes, if you tell an AI model it's wrong — even when it actually is wrong — it will insist it's right. Always verify.
AI isn't always great at math. Large language models were built for words, not numbers. Even though they've improved, mathematical reasoning isn't their core strength.
Knowing "just enough" AI is hard to judge. Where's the line between helpful support and replacing your own thinking? The only way to develop that judgment is to keep reflecting on it.
You don't have to get the prompt right the first time. AI conversations are iterative. Follow-up questions often produce better results than a single perfect prompt.
AI isn't a person — even though it sounds like one. It uses human language patterns convincingly, but there's no understanding, no experience, and no emotions behind it.
Transparency matters. If AI did a significant part of your work, others generally want to know. This is called disclosure, and it applies in school and in professional life.
Learning is still worth it — even when AI exists. You need to know things well enough to judge whether AI is right. Knowledge is still power; it just looks different now.
AI will likely touch every career. Not just tech jobs — agriculture, healthcare, marketing, education, and more. Understanding AI now is preparation for any future.
If you're not sure how AI can help, ask it. Tell the tool what you're trying to do and ask how it could assist. Then decide whether any of its suggestions are actually useful.
AI makes its best statistical guess. It's not looking up a fact or reasoning through logic — it's predicting the most likely next word or idea based on patterns. That's why it makes mistakes.
"Be the Bot" Activity
Be the Bot is Miller's signature classroom strategy — and it works in any subject. Students predict how AI will respond to a subjective question about what they're studying, then compare their thinking to the actual AI response and critique both. It builds content knowledge and AI literacy simultaneously.
Ask a good question
Choose a subjective, judgment-based question — not one with a clear factual answer. You want something that requires weighing evidence and making a case. Students will predict AI's answer, so the question needs to have enough complexity for meaningful comparison.
Question Starters That Work
Student brainstorm first
Students work in small groups to brainstorm and predict AI's response before seeing it. This is retrieval practice — recalling what they know strengthens long-term memory. Then they rank their predictions and justify them with evidence, which pushes into higher-order thinking.
The big reveal and class critique
Type the question into the AI tool and share it with the class. As results appear, students naturally compare, react, and debate. Then structure the critique: What did AI get right? What's missing? What biases can you spot? What did its response leave out?
Subject-Specific Examples
Who is the most courageous character in this novel?
Students analyze character traits and defend their ranking before comparing to AI's take.
Who were the three most pivotal figures in [historical event]?
Students apply historical knowledge to rank figures, then critique AI's reasoning for inclusions and omissions.
Which of these five environmental problems is most urgent to solve?
Students use scientific evidence to prioritize, then compare their reasoning to AI's and question its criteria.
What are the most useful types of math in everyday life?
Students defend their choices, then critique whether AI's answer reflects bias toward certain careers or contexts.
What is the most important habit for a healthy lifestyle?
Reveals AI's tendency to reflect Western health assumptions — a rich discussion point.
Who/what most belongs in this category we've studied?
Flexible enough for any unit — ask AI to categorize, sort, or rank something from your curriculum.
After the reveal — questions to push the critique deeper
- Was the format of AI's response a ranked list or just a list? Did it choose that format well?
- Is anything missing or inaccurate? What perspective does AI seem to be taking?
- Did AI respond the way you predicted? What would a human expert have said differently?
- Can you spot a bias? What kind of data probably shaped this response?
The ABC Rule: Always Be Critiquing
Every time AI produces something — an image, a summary, a recommendation, an explanation — it made dozens of invisible judgment calls you never saw. Critiquing AI output is a habit, not a one-time activity. The more students do it, the sharper their thinking becomes — about AI and about everything else.
AI-Generated Images
Spot artifacts, count fingers, check proportions — find the "AI weirdness"
AI-Written Summaries
Check accuracy, identify what's missing, flag where perspective is skewed
AI Recommendations
Notice what it defaults to — and what it systematically overlooks
Critique Frameworks to Use with Students
Start with straightforward questions
Before diving deep, open with low-stakes observation questions to get students engaged and noticing:
"What do you think — is this effective?"
"Can you tell it was made by AI? How?"
"Is there anything that's just 'not right'?"
"What do you see? What do you wonder?"
Note: You don't need to have the "right answer." Saying "I don't know either — let's investigate" is a powerful teaching moment.
Move from small details to big picture — or vice versa
When students get stuck at one level, prompt them to zoom in or out:
Put on someone else's shoes
Ask students to evaluate AI output from a different perspective — someone from a different background, time period, career, or part of the world. Ask students: "Whose worldview does this AI response best reflect? Whose voice is missing?"
When AI makes a mistake — lean into it
Imperfect AI outputs often make the best teaching moments. When AI produces something clunky, biased, or wrong, students immediately engage: "Wait, that's not right!" That reaction is the entry point. Don't rush past it — explore it. A flawed AI response can teach more about critical thinking than a perfect one.
15 Ethics Lenses for Any Lesson
These lenses can be applied to almost any topic students are studying to create rich discussion — about your content, about AI, and about how the two intersect. Each includes an AI-specific angle, but the lens works with or without AI in the conversation.
🎯 Integrity
How do we handle this responsibly? AI angle: At what point does AI assistance stop being "your own work"?
⚖️ Fairness
Is this cheating? Does anyone gain an unfair advantage? AI angle: When AI algorithms favor certain voices or outcomes, when does it become unfair?
🔧 Practicality
Could this be more helpful? Is it actually doable? AI angle: Which AI promises are practical vs. just hype?
🤝 Humanity
Does human involvement make the outcome better or worse? AI angle: What human connection is lost when we turn to AI for support?
📚 Learning
What do humans gain from completing this themselves? AI angle: Does AI use deepen understanding, or bypass the struggle needed for growth?
🔒 Safety
Could this put anyone at risk? How big is the risk? AI angle: How do we protect digital and physical safety in AI-mediated situations?
✅ Appropriateness
Does this behavior meet civil standards? AI angle: If we incorporate AI here, are we crossing social or moral boundaries?
👥 Perspective
Is one viewpoint overrepresented? Whose is missing? AI angle: Which voices has AI amplified — and which has it minimized?
🔍 Transparency
Is the process clear? What's hidden? AI angle: Should we disclose AI use — and should we know how AI reached its conclusions?
🎨 Creativity
Where does this creativity come from? Did external help diminish it? AI angle: How much AI assistance before a work is no longer "human creativity"?
🎮 Autonomy
Who or what is in control? Is there manipulation behind the scenes? AI angle: How does AI shape our choices and agency — who really holds the power?
🤔 Trust
Is trust based on evidence, or implicit assumption? AI angle: How much trust do AI systems actually deserve — and why?
⚡ Efficiency
Does improving efficiency cost us something important? AI angle: What do we gain — or lose — when we turn to AI for efficiency?
✍️ Consent
Do participants understand the implications? AI angle: What does consent mean in a digital world — data collection, facial recognition, AI art training?
💛 Empathy
Are we accurately considering others' emotions? What assumptions might be wrong? AI angle: How should we use technology that can only simulate, not feel, human emotion?
Building a Culture of Academic Integrity
Blocking and detection aren't solutions — they're games students learn to beat. Miller's approach shifts from policing to co-creation: invite students into an honest conversation about AI's role in their learning, and let them help set the norms. Note: for JA's specific guidance on AI detection tools and suspected misuse, see the Detecting AI Writing and Suspected Inappropriate Use pages.
Start with AI in professional life — not classwork
Before talking about AI use in your class, discuss how professionals in various fields are navigating the same tension. Students judge these situations more impartially when their own grades aren't on the line.
Discuss AI in your specific academic tasks
After the professional conversation, transition to classwork. For each type of assignment, talk through both ends of the spectrum — total AI use to no AI use — and explore where the line should be.
Co-create shared class norms
Rather than handing students a list of rules, use questions to help them build their own. When students help define the norms, they're more likely to hold themselves and each other accountable.
Discussion Questions for Co-Creating Norms
- How have you seen AI impact classwork — in helpful and harmful ways?
- What are the real reasons students might overuse AI? (Not to judge — to understand)
- What can I do as your teacher to reduce temptation to use AI irresponsibly?
- What should you do if you're unsure whether a use of AI is OK?
After an assignment — reflection questions about AI use
A quick reflection after classwork is one of the most powerful ways to build students' AI judgment over time. Consider adding one or two of these to your exit tickets or assignment submissions:
"Did I use AI too much on this?"
"Did AI support my learning or stand in the way of it?"
"Am I a more capable person as a result of this work?"
"Did AI help me express my ideas — or replace my voice?"
"Am I proud of this work? Does my use of AI change that?"
Ready-to-Use AI Prompts
These prompts, adapted from the book's "Prompt and Ponder" sections, are designed to help you personally explore AI literacy concepts — and to generate classroom-specific ideas for your own content area. Paste them into Claude, Gemini, or ChatGPT.
Generate "By the Way" Lessons for your class
Design a "Be the Bot" activity for your unit
Find the right AI artifact to critique with your class
Build shared integrity norms with students
Connect your curriculum to AI ethics discussions
Quick Reference: The Core Toolkit
Everything on this page distills to four moves you can make in any class, any day.
💬 "By the Way" Lesson
30 seconds. One AI truth. Connect it to today's content.
🤖 Be the Bot
Students predict AI's answer first, then compare and critique.
🔍 Always Be Critiquing
Treat every AI output as an opportunity to examine what it got right, wrong, and biased.
🤝 Co-Create Norms
Involve students in defining where AI helps vs. where it gets in the way of learning.
Source: AI Literacy in Any Class, Matt Miller (Ditch That Textbook, 2026). This page summarizes key frameworks for professional educator use at Jefferson Academy Secondary.