On February 13, 2026, the U.S. Department of Labor released something that every K–12+ educator in the country should read: a national AI Literacy Framework. Published through Training and Employment Notice 07-25, the framework lays out five foundational content areas and seven delivery principles designed to guide AI literacy efforts across the entire workforce and education system.
This isn’t a mandate. There’s no new regulation, no required curriculum, no compliance deadline. It’s a framework—voluntary guidance that gives schools, workforce boards, and training providers a shared vocabulary and structure for teaching AI literacy. The federal government is now on record saying that every American worker needs AI literacy, and it has spelled out what that means in concrete terms.
If you’re a K–12+ educator, this is your signal. Not because Washington is forcing your hand, but because the framework articulates what many of us already know: students who graduate without understanding how to use, evaluate, and think critically about AI tools will be at a serious disadvantage. The framework gives you a structure to act on that knowledge. Here’s what’s in it, and what you should do about it.
And let me say this plainly: when it comes to AI literacy, it will be the institutions that lead—not those that follow—that actually prepare their students. The schools, nonprofits, and community organizations that move first, experiment early, and build AI literacy into their programs now will produce graduates and participants who are ready for the economy that’s arriving. The ones that wait for perfect curricula, complete consensus, or top-down mandates will find themselves playing catch-up while their students pay the price. There is no reward for caution here. The only advantage belongs to those who act.
What the Framework Actually Says
The framework identifies five foundational content areas—the core of what any AI literacy effort should cover:
Understand AI Principles. Students need a working mental model of what AI is and how it operates. Not at the level of an engineer, but enough to use AI tools with confidence and good judgment. This means understanding that AI produces probabilistic outputs rather than definitive answers, that it can hallucinate, that it works across modalities like text, images, and audio, and that every AI system reflects human design choices about data, goals, and parameters.
Explore AI Uses. Students need exposure to how AI is actually being used in workplaces—drafting documents, analyzing data, generating creative assets, supporting decision-making, automating routine tasks. Because AI use varies enormously by industry and role, exploration builds the familiarity students need to recognize when AI can help and when human judgment is more appropriate.
Direct AI Effectively. This is the art of prompting. The framework treats it not as a technical skill but as a communication skill applied to a new kind of tool. Students need to understand contextual framing, clear instruction structure, iteration, and the importance of supplying relevant examples and data.
Evaluate AI Outputs. AI can accelerate work, but its outputs require human review. Students should learn to verify factual accuracy, assess completeness, spot logical gaps, and apply their own expertise to decide how to use, revise, or discard what AI produces. The core principle: workers must remain in control.
Use AI Responsibly. This covers protecting sensitive information, following workplace policies, avoiding misuse such as plagiarism or impersonation, managing risk in high-stakes settings, and maintaining personal accountability for decisions made with AI assistance.
How to Teach It: The Seven Delivery Principles
The framework doesn’t just say what to teach. It says how. Seven delivery principles guide program design:
Enable experiential learning. People learn AI by doing, not by reading about it. Programs should embed hands-on practice with real AI tools into real tasks—writing, research, analysis—with progressive difficulty and live feedback.
Embed learning in context. AI literacy sticks when it’s tied to a learner’s actual work or field of study. This means using industry-specific examples, aligning content with occupational tasks, and integrating AI literacy into existing CTE curricula and apprenticeship programs.
Build complementary human skills. This is the principle I want to spend the most time on, so I’ll unpack it in the next section.
The context starts at the top. In a recent HBR interview, McKinsey’s global managing partner Bob Sternfels offered a striking formulation: “When people ask me how many people McKinsey employs, my answer is 60,000: 40,000 humans and 20,000 agents.” Eighteen months ago, the firm had 3,000 agents….That reorganization is changing what McKinsey values in a hire. Sternfels named three things AI can’t do: set the right level of aspiration, exercise judgment, and make discontinuous leaps in thinking. The firm is recruiting more from liberal arts backgrounds it had “deprioritized” and looking for “aptitude to learn new stuff” over mastery of specific subjects. (Kabir)
Address prerequisites. Ensure learners have the foundational digital literacy, device access, and connectivity needed to participate. Not everyone starts from the same place.
Create pathways for continued learning. Foundational literacy is just a starting point. Programs should build stackable progressions toward deeper AI proficiency and even AI builder and entrepreneurship pathways.
Prepare enabling roles. Managers, trainers, mentors, and counselors need their own AI literacy so they can support others. Train-the-trainer models and peer learning champions are essential.
Design for agility. AI evolves faster than traditional curricula. Build in modular content, feedback-driven revision, and regular updates so your program doesn’t become obsolete before it launches.
The Most Important Idea in the Framework: Build Complementary Human Skills
Delivery principle number three deserves its own section because it carries the most important insight in the entire document: AI tools are amplifiers of human input. The quality of what AI produces depends heavily on the skills, knowledge, and judgment of the person using it. A student with strong critical thinking will get better AI outputs, evaluate them more effectively, and apply them more wisely than one without those skills. AI doesn’t replace the need for human capabilities. It makes them more consequential.
The framework identifies five areas of complementary human skill that should be developed alongside AI literacy:
Critical thinking. When students use AI to solve problems, the learning experience should reinforce human judgment as central to AI-supported decisions. Students shouldn’t just accept an AI output—they should interrogate it, test its assumptions, and decide whether it actually solves the problem. In practice, this might mean giving students an AI-generated analysis of a historical event and asking them to identify what the AI got wrong, what it oversimplified, and what perspectives it missed.
Creativity. AI can brainstorm, generate variations, and remix ideas—but the human’s role is curatorial and generative. Students should use AI to expand their raw material and then apply their own aesthetic judgment, purpose, and originality to shape the result. A design student might use AI to generate twenty concepts and then apply their own judgment to select and refine the best direction.
Communication. AI can draft content, but humans need to revise it for tone, clarity, persuasiveness, and audience awareness. Communication actually becomes more important in an AI context, because workers now need to communicate effectively both to AI (through good prompts) and to other humans (by refining what AI produces). A writing teacher might use AI to generate first drafts and then focus classroom time on the distinctly human work of revision.
Values-based decision-making. AI can surface options and data, but navigating ambiguous situations where organizational, legal, or personal values are at stake requires human judgment. A business ethics course might present a scenario where AI recommends a cost-cutting measure that’s technically legal but ethically questionable, and students must reason through the decision themselves.
Domain expertise. The value of AI increases dramatically when the person using it brings deep subject-matter knowledge.
The practical takeaway is that AI literacy should not be taught as a purely technical skill divorced from the rest of the curriculum. The most effective approach is to integrate AI tools into existing courses in ways that simultaneously develop students’ critical thinking, creativity, communication, ethical reasoning, and domain knowledge—because these are the skills that make AI useful rather than misleading, and that make people valuable rather than replaceable.
Beyond Literacy: Developing Deeper AI Skills
The framework recognizes that AI literacy is the floor, not the ceiling. Schools should also be developing skills that sit above and around basic literacy.
Data literacy—understanding how data is collected, structured, analyzed, and interpreted—is a natural complement. Workers who understand data can better assess AI outputs, recognize bias in training data, and make more informed decisions about when to trust AI recommendations.
The framework envisions a progression from foundational literacy through AI proficiency (managing more complex AI systems) to AI builder skills (designing, configuring, or building AI-powered solutions). Schools should create stackable learning models that allow students to move along this spectrum based on their career goals.
And the framework specifically calls for supporting pathways into entrepreneurship—students who want to go beyond using AI tools to building their own AI-powered solutions. Project-based learning, hackathons, and partnerships with local incubators or startup ecosystems can all foster this.
A Natural Home: Academic Competitions
If you’re looking for places where AI literacy and complementary human skills come together organically, look no further than competitive academic activities. Debate, Model UN, mock trial, speech and forensics, Science Olympiad, academic decathlon—these are environments where the framework’s vision of AI-augmented human capability plays out in real time.
Debate is perhaps the clearest example. AI literacy integrates at every stage of preparation: students can use AI to research arguments, generate counterarguments, stress-test their cases, and draft briefs. But every one of those uses demands the complementary human skills the framework emphasizes. Evaluating AI outputs is critical—debaters who trust AI-generated evidence without verifying it will get caught with fabricated citations or distorted claims. Directing AI effectively matters because a debater with deep topic knowledge will get far better research assistance than one typing vague prompts. And the skills that actually win rounds—critical thinking to weigh competing arguments on the fly, communication to persuade a judge, values-based reasoning to navigate complex policy trade-offs—are irreducibly human. AI accelerates preparation; human judgment drives performance.
Model UN follows a similar pattern. Delegates can use AI to research country positions, draft position papers, generate resolution language, and prepare for committee sessions. But the competitive differentiator is domain expertise—knowing the nuances of your assigned country’s foreign policy—combined with communication skills for real-time negotiation and creativity for proposing novel solutions under pressure. Responsible use matters here too: delegates need to recognize when AI-generated policy positions are plausible versus when they’re hallucinated nonsense about a country’s actual stance. The student who can direct AI with precision and then evaluate what comes back with genuine subject-matter knowledge will outperform the one who copies and pastes.
Mock trial, and other competitions all follow the same logic. AI can help a mock trial attorney draft an opening statement, but the persuasiveness of the delivery, the ability to adapt to unexpected witness testimony, and the ethical reasoning required to make judgment calls in real time are all human.
The broader point is this: academic competitions are already teaching the complementary human skills the DOL framework identifies. They just haven’t been framed that way. If you coach or advise any of these activities, you have a ready-made laboratory for AI literacy integration—one where students are intrinsically motivated to learn because winning depends on it.
The Call to Action: Don’t Wait
Here’s what I want every K–12 educator reading this to hear: you do not need to wait for your state, your district, or your administration to tell you to start. The framework is voluntary. It’s a resource, not a regulation. That means the schools that move first will be the ones that serve their students best.
Start by auditing your existing curriculum for natural integration points. AI literacy doesn’t require a new standalone course. The five content areas map naturally onto subjects you already teach: prompting is a communication skill that belongs in English class, evaluating outputs is a critical thinking exercise that fits in science or social studies, responsible use is an ethics conversation that can happen anywhere, and exploring AI uses is a natural extension of career and technical education.
Then prioritize hands-on experience. Give students access to AI tools and let them use those tools to complete real assignments. The framework is emphatic on this point: people learn AI by doing. Side-by-side comparisons of AI-generated and human-created work, exercises identifying hallucinations, and progressively complex prompting scenarios all build genuine capability.
Most importantly, don’t treat AI literacy as separate from the skills that matter most. The framework’s most powerful insight is that critical thinking, creativity, communication, ethical reasoning, and domain expertise aren’t competing with AI skills for classroom time—they’re the foundation that makes AI skills valuable. Teach them together.
The students sitting in your classrooms right now will enter a labor market that assumes AI literacy as a baseline. The Department of Labor has laid out a clear, well-structured framework for what that literacy looks like. The question isn’t whether to act. It’s how quickly you can start.
Over to You: What Is Your School or Organization Doing?
Now I want to turn this into a conversation that actually leads somewhere. Not “what do you think about AI?” but “what are you going to do on Monday morning?”
If you’re at a school—a challenge:
Your students are already using AI—at home, on their phones and in their glasses, for your assignments whether you know it or not. The only question is whether they’re learning to use it well or learning to use it badly. Which is your school producing right now?
Name one teacher in your building who could pilot an AI-integrated unit next quarter. Now ask yourself: what is stopping you from giving them the green light this week?
If you banned or restricted AI tools in your school, what was your plan for teaching students how to use them responsibly? If there was no plan, the ban wasn’t a strategy—it was an abdication. What’s your next move?
Pick one assignment you give regularly. How would it change if students were required to use AI as part of the process—and then required to explain where the AI helped, where it failed, and what they had to fix? Could you redesign that assignment by the end of this month?
The framework says critical thinking, creativity, communication, and ethical reasoning should be taught alongside AI—not in a separate unit, but woven into how students actually use the tools. Which of your existing courses is best positioned to do this tomorrow, with zero new budget and zero new curriculum? What’s the first step?
If you run a debate program, a Model UN team, a mock trial squad, or any academic competition—you already have a laboratory for this. Your students are already building the exact complementary human skills the framework calls for. Have you explicitly connected those activities to AI literacy yet? If not, that’s the lowest-hanging fruit in your entire school.
A year from now, when your graduates are sitting in college classrooms or job interviews where AI fluency is assumed, will they be ready? If the honest answer is no, what are you doing between now and then to change it?
If you’re at a nonprofit or community organization—you own this responsibility too.
Let me be direct about something: if your organization serves people—trains them, educates them, prepares them for employment, supports their development—then you have the same obligation to teach AI literacy that schools do. This is not a school problem. It is an everyone-who-touches-learners problem. The DOL framework was written for you just as much as it was written for K–12 systems. Your participants will enter the same AI-shaped labor market as every high school and college graduate. If they leave your programs without understanding how to use, evaluate, and think critically about AI, you have left them unprepared—full stop.
The difference is that you can act faster. Schools are tankers. You’re a speedboat. You don’t need a school board vote, a two-year curriculum review, or a state standards revision. You can stand up a pilot program next month, learn from it, and iterate while school systems are still scheduling their first planning meeting. Equal responsibility, greater agility. That combination should make you the first movers, not the last.
What would it take for you to run a four-week AI literacy pilot for the people you serve—not someday, but starting within 60 days? What’s actually in the way? Be specific.
The people you serve—job seekers, immigrants, formerly incarcerated individuals, workers in declining industries—are the ones the AI economy will hit first and hardest. Every month you wait to build AI literacy into your programming is a month your participants fall further behind people who already have access. What is your timeline?
Your local schools are struggling to figure out AI literacy on their own. Could you be the partner that provides the hands-on, experiential learning component they can’t build fast enough? An after-school AI workshop series, a summer intensive, a weekend bootcamp for students and parents together—you could launch something like this before most districts finish their first committee report. Have you pitched it?
Is your own staff AI-literate? If you’re going to teach this, your team needs to live it first. Are your program managers using AI to improve their own work? Are your trainers comfortable enough with the tools to teach others? If not, start there—internal capacity building is the prerequisite for everything else.
What’s the scrappiest, most inventive AI literacy effort you’ve seen or built? I’m not looking for programs with six-figure budgets. I’m looking for the organization that figured out how to teach prompting in a public library with donated laptops, or the workforce program that embedded AI skills into existing job readiness training without adding a single new line item. Those are the models that will actually scale. Share them.
Tell me what you’re going to do. Not what you think about AI literacy. Not what you wish your organization would prioritize. What you are personally going to set in motion this month. The comments section isn’t for reflection—it’s for commitments.