Plot Twist: Teachers Have Been Teaching AI Literacy Since Before AI Existed
- Michelle Strom
- Jul 16
- 5 min read
There’s a lot of noise right now about AI literacy—what it is, where it goes, who’s responsible for it, whether schools are doing enough, and whether teachers are prepared. If you’ve sat through a professional development session about it recently, you may have left feeling like there’s an entirely new body of knowledge to absorb before you’re allowed to call yourself ready.
That framing gets something important wrong. AI literacy (the real kind, not the “let’s teach kids to use chatbots” kind) is really about one thing: teaching students how to think when something is producing information for them. How to ask whether it’s accurate, decide whether to trust it, question it, or set it aside entirely.
Sound familiar? It should, because that’s what good teachers have been doing since long before anyone had heard of a large language model.
The assumption buried in most AI literacy discussions
One reason this realization matters is that most AI literacy conversations start from the idea that it’s a response to something new. But it isn’t, really.
The scientific method was already asking the right questions. And every English teacher who has ever said “but what is the author trying to do here?” was doing AI literacy before it had a name. Same with science teachers who were pushing students to question methods and evidence, Social studies teachers asking whose voices were included and whose were missing, and math teachers reminding students that probability isn’t certainty.
But what AI did was raise the stakes on all of it.
Students can now generate essays, images, explanations, and answers in seconds, and those outputs often sound authoritative even when they’re incomplete, misleading, or wrong.
And that comes from better judgment: knowing when to trust, when to question, when to dig deeper. And those aren’t computer science skills, they’re thinking skills: the kind teachers have been building all along.
How AI literacy shows up in every subject
The ELA teacher who reads between the lines
Let's start with what's already happening in English class.
Every time an ELA teacher asks a student to evaluate a source, they lay the groundwork for AI literacy. Every time a teacher helps a student understand that something can be written with total confidence and still be completely wrong, they build the muscle students need to read AI outputs critically.
And today’s version of that challenge is weirder: large language models generate text that sounds authoritative without any real human reasoning behind it. The AI didn’t have a point of view, but it produced the shape of one. A student who has been taught to read for what a text is doing, not just what it says, has a real shot at noticing that and at asking: who is speaking here? What are they trying to get me to believe? What’s missing?
Those are ELA questions. They have been since long before AI generated a single sentence.
The science teacher who pushes for proof
The same pattern holds in science, but with different vocabulary.
The scientific method is built on the premise that you can’t just believe something because the results look good. You have to ask how the data was collected, consider what might have been missed, and think about what would have to be true for this result to be wrong.
Science teachers have spent their careers building that habit in students—and it’s what keeps a student from taking an AI output at face value.
The social studies teacher who challenges the narrative
If science asks students to question data, social studies asks them to question something harder: the story itself.
History is full of confident, biased sources that overlook critical voices. Social studies teachers have always pushed students to ask: Who wrote this? What did they have to gain? Whose story was left out of the record?
Those questions have been the engine of good history and civics instruction for as long as there has been history and civics instruction, and they’re exactly the questions students need to ask about AI systems.
The math teacher who questions the odds
And then there’s math that asks the question that ties all of it together: where does the answer even come from? Because AI doesn’t actually know things—it just predicts them.
When a large language model produces an answer, it isn’t retrieving a fact from a database. It’s calculating the most statistically likely next word, and the one after that, and the one after that, based on patterns in everything it was trained on.
While the output looks like knowledge, underneath, it’s probability. And when students understand that, they’re less likely to treat AI outputs as settled fact.
It's not just in the curriculum. It's in the conversations.
What makes this especially interesting is that the work isn't confined to any single subject. It's happening everywhere, students and adults are thinking out loud together:
The art teacher who pauses class when a student traces someone else’s drawing and calls it their own, to discuss where inspiration ends, copying begins, and who owns an idea.
The librarian who watches a student accept the first search result without looking further and says, gently, “But did you check where that came from?”
The school counselor who sits with a student in the middle of a rumor spiral and asks, before anything else: How do you know this is true? Who told you? What’s their relationship to the situation?
The kid on the bus who holds up their phone and says, “Wait, is this video even real?” and the other kid who says, “I don’t know, how would we even know?”
AI made the cost of not teaching these habits visible in a way that wasn't as urgent before. Every teacher who has ever pushed a student to think harder before accepting something at face value has already been doing the work.
What makes teachers well-suited for this
Which brings us to what AI literacy actually asks of teachers (and what it doesn’t.)
It doesn’t require a computer science degree, a crash course in machine learning, or an understanding of how large language models are trained.
It does require the ability to ask good questions and model what it looks like to reason through uncertainty out loud—to sit with a student in front of an AI output and say: "I don't know if this is right. Here's how I'd think about whether to trust it."
And that kind of transparent, humble reasoning—thinking out loud, naming the uncertainty, showing the process—is one of the most powerful things a teacher can do. It's also exactly what great teachers have always done.
How the BrainPOP AI Literacy Collection supports this work
This is also the premise the BrainPOP AI Literacy Collection was built on: that AI literacy is embedded in the curriculum teachers are already responsible for, not added on top of it.
Researchers who study this agree. The AI4K12 initiative's Five Big Ideas in AI map directly onto concepts already present in core subject areas, and the work BrainPOP co-created with Digital Promise makes the same point. Students don't need a dedicated AI course to develop AI literacy; they need teachers who can make the connections between what students already learn and how AI systems work explicit.
Each topic in the collection connects to a conversation that's already happening in your classroom. Not a separate unit, but a deeper layer than the one you're already teaching.
Michelle Strom is Associate Director, Product Marketing at BrainPOP. She holds a Master’s in Educational Psychology and is committed to keeping learners at the center of how products are built, marketed, and understood. She is a fan of procedural crime dramas and carrot cake. Her favorite BrainPOP character is Nat.