Everywhere you look, someone is telling students and workers to “learn AI.”
It’s become the go-to advice for staying employable, relevant and prepared for the future. But here’s the problem: While definitions of artificial intelligence literacy are starting to emerge, we still lack a consistent, measurable framework to know whether someone is truly ready to use AI effectively and responsibly.
And that is becoming a serious issue for education and workforce systems already being reshaped by AI. Schools and colleges are redesigning their entire curriculums. Companies are rewriting job descriptions. States are launching AI-focused initiatives.
Yet we’re missing a foundational step: agreeing not only on what we mean by AI literacy, but on how we assess it in practice.
Two major recent developments underscore why this step matters, and why it is important that we find a way to take it before urging students to use AI. First, the U.S. Department of Education released its proposed priorities for advancing AI in education, guidance that will ultimately shape how federal grants will support K-12 and higher education. For the first time, we now have a proposed federal definition of AI literacy: the technical knowledge, durable skills and future-ready attitudes required to thrive in a world influenced by AI. Such literacy will enable learners to engage and create with, manage and design AI, while critically evaluating its benefits, risks and implications.
Second, we now have the White House’s American AI Action Plan, a broader national strategy aimed at strengthening the country’s leadership in artificial intelligence. Education and workforce development are central to the plan.
Related: A lot goes on in classrooms from kindergarten to high school. Keep up with our free weekly newsletter on K-12 education.
What both efforts share is a recognition that AI is not just a technological shift, it’s a human one. In many ways, the most important AI literacy skills are not about AI itself, but about the human capacities needed to use AI wisely.
Sadly, the consequences of shallow AI education are already visible in workplaces. Some 55 percent of managers believe their employees are AI-proficient, while only 43 percent of employees share that confidence, according to the 2025 ETS Human Progress Report.
One can say that the same perception gap exists between school administrators and teachers. The disconnect creates risks for organizations and reveals how assumptions about AI literacy can diverge sharply from reality.
But if we’re going to build AI literacy into every level of learning, we have to ask the harder question: How do we both determine when someone is truly AI literate and assess it in ways that are fair, useful and scalable?
AI literacy may be new, but we don’t have to start from scratch to measure it. We’ve tackled challenges like this before, moving beyond check-the-box tests in digital literacy to capture deeper, real-world skills. Building on those lessons will help define and measure this next evolution of 21st-century skills.
Right now, we often treat AI literacy as a binary: You either “have it” or you don’t. But real AI literacy and readiness is more nuanced. It includes understanding how AI works, being able to use it effectively in real-world settings and knowing when to trust it. It includes writing effective prompts, spotting bias, asking hard questions and applying judgment.
This isn’t just about teaching coding or issuing a certificate. It’s about making sure that students, educators and workers can collaborate in and navigate a world in which AI is increasingly involved in how we learn, hire, communicate and make decisions.
Without a way to measure AI literacy, we can’t identify who needs support. We can’t track progress. And we risk letting a new kind of unfairness take root, in which some communities build real capacity with AI and others are left with shallow exposure and no feedback.
Related: To employers, AI skills aren’t just for tech majors anymore
What can education leaders do right now to address this issue? I have a few ideas.
First, we need a working definition of AI literacy that goes beyond tool usage. The Department of Education’s proposed definition is a good start, combining technical fluency, applied reasoning and ethical awareness.
Second, assessments of AI literacy should be integrated into curriculum design. Schools and colleges incorporating AI into coursework need clear definitions of proficiency. TeachAI’s AI Literacy Framework for Primary and Secondary Education is a great resource.
Third, AI proficiency must be defined and measured consistently, or we risk a mismatched state of literacy. Without consistent measurements and standards, one district may see AI literacy as just using ChatGPT, while another defines it far more broadly, leaving students unevenly ready for the next generation of jobs.
To prepare for an AI-driven future, defining and measuring AI literacy must be a priority. Every student will be graduating into a world in which AI literacy is essential. Human resources leaders confirmed in the 2025 ETS Human Progress Report that the No. 1 skill employers are demanding today is AI literacy. Without measurement, we risk building the future on assumptions, not readiness.
And that’s too shaky a foundation for the stakes ahead.
Amit Sevak is CEO of ETS, the largest private educational assessment organization in the world.
Contact the opinion editor at [email protected].
This story about AI literacy was produced by The Hechinger Report, a nonprofit, independent news organization focused on inequality and innovation in education. Sign up for Hechinger’s weekly newsletter.
