AI literacy is present in every K-12 classroom

Students interact with AI faster than schools can respond. Districts need a real literacy strategy, not just a usage policy.

New York City gave 600,000 elementary and middle school students one answer to the artificial intelligence question, and it’s no. The nation’s largest school district announced it a one-year moratorium on generative artificial intelligence aimed at students through eighth grade, calling it the “widest ban of its kind in the country.” High school students get limited access and mandatory “critical thinking” modules on the risks of relying on technology.

It’s a reasonable response to a real problem. But it also illustrates the challenge schools face around the world: Students encounter artificial intelligence whether or not it has been formally introduced into the classroom, and districts and teachers need a plan to help them understand how to handle it responsibly.

The gap between faith and practice

Most teachers already agree that AI needs a place in education. In an Ipsos survey conducted earlier this year, 78% of teachers said responsible use of AI should be part of their school’s curriculum. But 52% said their school had provided no guidance on what this looked like or were unsure whether guidance existed or not.

The political picture appears similar. A 2025 Milken Institute report found that most US schools still lack AI education standards or relevant teacher skills, warning that uneven implementation is becoming a real risk as AI use grows.

Students also feel this gap. According to a 2025 study, more than half of students use AI for schoolwork RAND studybut only 35% of district leaders said they provide students with AI training. Over 80% of students said their teachers did not explicitly teach them how to use AI in school.

Treat AI literacy as a life skill

Schools have a long history of teaching students how to deal with the real-world situations they will face, regardless of their career path. Financial literacy, interviewing skills, digital citizenship – these are taught because schools recognize that some skills are too consequential to be left to chance and too broadly applicable to fit into a single subject area.

AI literacy belongs in that same category. It is a computer science topic in the same limited sense that budgeting is a math topic. Treating it this way limits it to students who choose to take programming classes. Students will encounter AI at school, at work, and in everyday life, often through technologies they may not even recognize as AI. Preparing them means teaching not only how to use a particular tool, but how artificial intelligence works, where it shows up and how to evaluate what it produces.

AI literacy does not require every student to become an AI expert. Students need sufficient understanding to make informed decisions about the technology that will shape their lives.

Two skill sets, taught together

Responsible use of AI actually requires two distinct skill sets, and schools that focus on just one are preparing students for a false sense of competence.

The first is technical: understanding how AI tools generate responses, learning to verify results instead of accepting them at face value, recognizing biases and hallucinations, evaluating which tools are appropriate for a particular task (if any), and understanding what happens to personal data once it is entered into a prompt.

The second is human: critical thinking, empathy, adaptability, decision making and judgment. These are the skills that help students decide whether to use artificial intelligence, maintain their ability to learn and think for themselves, and build and sustain real relationships with the people around them. These are the skills most at risk when students rely on AI as a shortcut. For example, half of the high school students surveyed by the Center for Democracy and Technology they said that using AI in the classroom makes them feel less connected to their teacher.

Another 2025 study found that frequent use of AI tools was linked to weaker critical thinking performance, driven largely by a pattern the researchers call cognitive offloading, in which people delegate thinking to the tool instead of doing it themselves. The effect was more pronounced among younger participants.

The stakes go far beyond the classroom. The most recent from the OECD Survey on adult skills found that adult literacy and numeracy skills have declined or remained stagnant in most member countries over the past decade, and that adults with stronger numeracy skills are more likely to be employed and earn higher wages. Students who routinely entrust their thinking to artificial intelligence risk accelerating this trend, impacting their future earnings and the economy at large.

A student who can write a flawless suggestion but cannot evaluate whether the output is actually correct has not learned AI literacy. They have learned to trust a tool without understanding it. The true measure is whether the student can recognize when it is wrong or inappropriate, know when human judgment needs to take over, and continue to strengthen the skills and attributes that make us uniquely human.

How it looks in practice

To achieve this, a single radical mandate is not necessary. It requires some consistent efforts across the district.

First, instruction must be tailored to grade level. What a third grade student needs to understand about artificial intelligence is different than what a high school student entering college needs to understand, and a one-size-fits-all approach will fail both ends of that range. Responsible use must be developmentally appropriate, starting with basic awareness and digital safety toward more sophisticated issues of accuracy, bias, ethics, and appropriate use as students mature. It also works best when inserted into courses that students already take, such as career preparation, health, or communications, rather than inserted as a stand-alone unit that is covered once and forgotten.

Secondly, teachers need training before students. Asking educators to lead the responsible use of AI without first equipping them with an understanding of the tools puts them in an impossible position. Training should also provide educators and administrators with shared resources to conduct guided conversations with students, rather than letting each teacher set expectations independently.

So far, federal guidance hasn’t helped. Testimony at a recent U.S. House hearing described federal AI regulation for schools as effectively nonexistent, leaving districts to manage student safety, data privacy and staff training largely on their own. Districts cannot afford to wait for guidance that may not come, and they should not be forced to do so alone. State and national associations and other nongovernmental entities representing districts and educators are well positioned to develop model policies, convene communities of practice, and share what works.

Third, durable human skills need a dedicated space. As AI increasingly handles tasks that were once signature learning moments and critical responsibilities for entry-level workers, the skills that enable people to lead, exercise judgment, overcome ambiguity, and work effectively with others become more important, not less important. Employers see it too. In May 2026 Survey by the Strada Institute for the Future of Work Of nearly 1,500 talent leaders, a third said AI has reduced the critical skill-building tasks available to entry-level employees. Those same leaders ranked critical thinking and communication as the most important skills for new graduates, and AI literacy as the least important. AI skills still matter. More than a third of entry-level jobs now require them, nearly triple the share compared to fall 2025, according to the report. National Association of Colleges and Employers. But employers increasingly view AI-related skills as the foundation and durable skills as the differentiator. Young people entering that workforce will need more than just knowledge of tools to stay relevant and thrive. They will need judgment to manage, apply, limit and safeguard AI at work, as well as to collaborate and positively influence the people around them. Schools have the opportunity to develop these skills long before students enter the workforce, giving them repeated opportunities to solve problems, work with others, exercise judgment, and take responsibility for their own decisions.

The clock is already running

Students don’t have the luxury of waiting for schools to catch up. They are forming habits about how they use AI, for better or worse, and those habits will follow them into college and into their first job.

The goal should be to teach students how to use AI deliberately – when to question it and when not to use it at all. Schools have always played a role in preparing students for a world that extends beyond the classroom. AI literacy is now part of this responsibility.

Much of today’s debate about artificial intelligence focuses on what it might take away from us. Schools can prepare a generation that knows how to shape and apply this technology in ways that amplify our humanity rather than threaten it. Students will need the technical skills to work effectively with these tools, but they will also need the judgment, perspective, and human connection that technology cannot provide on its own. Preparing students to develop such capabilities alongside AI proficiency is what being ready for an AI-enabled future actually means.