The US Government Just Defined AI Literacy. Most Teachers Don’t Know It Happened.
As the “Make America AI Ready” initiative sets a new national standard, teachers must decide if workforce readiness is the same thing as true literacy.
This post accompanies the latest episode of What Teachers Have to Say called “Make America AI Ready: The Stove Isn't Going to Blow Up”. Listen wherever you get your podcasts or a whateteachershavetosay.com
Friday morning. 11:33 a.m. A text arrives.
Make America AI-Ready: Hey there! Noticed you haven’t responded yet. A quick reply, and we’re back on track. Did we catch you at a bad time?
The federal government, checking in on my AI education like Duolingo nudging me back to an abandoned lesson.
I had signed up a few days earlier: I texted READY to 20202, and just like that, the U.S. Department of Labor started delivering a free, seven-day AI literacy course directly to my phone. No app. No login. No cost. Ten minutes a day, one lesson at a time, in the same place I get texts from my kids’ school and the occasional spam about car warranties, an unsolicited loan offer, or a wayward Nigerian prince.
I want to be clear: I liked the program. The content was solid. It aligned closely with how I run trainings. The prompting structure (Goal, Context, Expectations) was clean and teachable. The garage organization prompt was genuinely clever: Snap a photo of your closet and ask AI to review it like a disappointed interior designer. There’s real wit in that. For someone who has never once opened ChatGPT, this is a good on-ramp; not a masterclass, a bite-sized trailer. Something that moves a person from not knowing what they don’t know to knowing, more precisely, what they don’t know yet.
That’s a legitimate and undervalued goal: Dunning-Kruger, meet public infrastructure.
But here’s what I’ve been sitting with since experiencing the program: the moment the Department of Labor launched their guidance and Make America AI Ready, a definition got institutionalized, and most teachers (the people who will be asked to build AI literacy in students) weren’t in the room when it happened.
The DOL defines AI literacy as a foundational set of competencies that enable individuals to use and evaluate AI technologies responsibly, with a primary focus on generative AI, which is increasingly central to the modern workplace.
Read that again. Workplace. Generative AI. Responsible use. The framework is built around five content areas: understanding AI principles, exploring uses, directing AI effectively, evaluating outputs, and using AI responsibly. These are reasonable things to know. The delivery principles are thoughtful: experiential learning, contextual embedding, and pathways to continued learning.
And then there’s the container it all sits inside. The DOL framework explicitly supports what the White House’s March 2026 National Policy Framework for Artificial Intelligence calls the administration’s reindustrialization agenda. It is designed to prepare American workers for an AI-driven economy. That is the frame. Workforce readiness. Employability. Competitive advantage.
That is one legitimate definition of AI literacy, but it’s not the only one. And it is not, by itself, the right one for schools.
Here is how definitions travel.
The White House establishes a policy direction. The Department of Labor builds a framework and a program that operationalizes it. State education agencies (thirty-one of them have now published AI guidance as of early 2026) read federal frameworks and, especially when/if funding is attached, orient toward them. State guidance shapes what county and regional offices of education like mine, and local districts work with. Those offices shape district decisions. Districts shape site choices. And somewhere at the end of that chain, a teacher gets told that AI literacy is something they’ll need to embed into their existing courses next year.
At each layer, nuance compresses. Compliance pressure accumulates. The original frame, workforce readiness and reindustrialization, doesn’t disappear. It just becomes invisible, because it’s now the compliance water everyone is swimming in.
California’s guidance (and others), to its credit, pushes toward something broader. It defines AI literacy per Education Code as the knowledge, skills, and attitudes associated with how artificial intelligence works: its principles, concepts, and applications, as well as its limitations, implications, and ethical considerations. It calls for AI literacy embedded across content areas, starting in elementary school and growing in sophistication through grade 12, oriented around the AI4K12 framework’s five big ideas: perception, representation and reasoning, learning, natural interaction, and societal impact. That last one (societal impact) is doing work that the DOL framework doesn’t ask of its learners at all. Like most states, California’s guidance is not mandatory yet. But guidance becomes practice, and practice becomes expectation, and expectation becomes policy, often before anyone votes on it.
Meanwhile, a teacher in rural Iowa or the Eastern Sierras (and if you’re reading this, statistically, you might be) is about to be handed a curriculum directive that was shaped, several layers upstream, by a definition they never encountered and a debate they were never invited into.
That is not a conspiracy, it’s just how policy development works. But knowing it happened is different from NOT knowing it happened.
There’s a second problem that the definitional question tends to obscure.
AI literacy is being rolled out to educators who, in many cases, haven’t experienced or learned it themselves. That’s a pretty important training problem, but it’s solvable; it just requires time, resources, and honest PD that meets people where they are. But the more complicated issue is this: some teachers will receive an AI literacy mandate and find that the underlying definition conflicts not just with their skill level, but with their core values and perceptions. A workforce-readiness frame assumes that equipping students to be productive workers is the goal. But, a teacher whose north star is to build better humans, someone who believes that education is about developing judgment, empathy, agency, and citizenship, may find themselves trying to teach toward a definition that doesn’t quite capture what they’re actually trying to do. Personally, I like the idea of developing rad humans.
Lack of competence is a training problem, and ideological difference is a values problem; they require different responses, and no one in the transmission chain right now is distinguishing between them yet, at least that I’ve seen.
So what can a teacher actually do?
Start here (assuming you’ve played with an LLM before. If not, stop reading and try one out). Before any adopted framework or district initiative roll-out, before any curriculum mandate, before the professional development session where someone walks you through the DOL’s five content areas: ask a student to prompt an AI image generator with one word. Padlet has an easy option, if you’re looking for one.
Prompt: Make me an image of [choose one: Shoes, Scientist, Doctor, Family, any simple noun will do.]
Watch what comes back. Then ask: what are we seeing? who made these choices? Whose experience is centered? What’s missing? What would happen if we changed the prompt? Then, have students generate another image correcting for the bias.
That’s the activity at the heart of Shoe Bias, one of the EduProtocols in the forthcoming EduProtocols AI Literacy Edition by Kate Meyer and Nicole Davis, two Maine teachers who have spent the past few years building and testing AI literacy frameworks in real classrooms. The HUMAN framework at the core of their book (Halt, Utilize, Monitor, Authenticate, Note) is built around a different question than the DOL framework asks. Not “can you use this tool productively?” but “can you stay in the driver’s seat and think critically about what you’re getting back?”
That distinction is not trivial. The Shoe Bias EduProtocol isn’t a workforce readiness activity, but a critical thinking activity that happens to use AI as the material. It develops exactly the kind of reasoning the DOL framework’s “responsible use” section gestures toward but doesn’t quite reach: not just protect sensitive information and follow workplace policies, or organize your garage, but notice what the system assumes, name what it leaves out, and understand that you are the one who decides what to do with that.
You can preview the book at tinyurl.com/epaipreview.
Somewhere in the next eighteen months, I predict that most schools will begin rolling out something they’ll call AI literacy. The definition they’re likely going to be working from has it’s origins in what applies to workers. That doesn’t make it wrong, but it likely makes it incomplete for our students.
The teachers who will do this well are the ones who already know the difference between a floor and a curriculum; they will take whatever framework arrives in their inbox and ask: what does this assume, what does it leave out, and what does my version of this need to include in order to serve the students in front of me?
That question has always been the job. AI literacy doesn’t change it. It just makes it more urgent.
This episode of What Teachers Have to Say goes deeper on the Make America AI Ready program, the DOL framework, and what AI literacy needs to look like for teachers and students. Listen wherever you get your podcasts or visit: whatteachershavetosay.com