The U.S. surgeon general’s advisory urging schools to limit screen time, invest in physical textbooks, and return to paper-and-pencil instruction arrives at a pivotal, and deeply contradictory, moment in American education.
Students will need to know how to use AI productively if they hope to succeed in the job market, but new bans and regressive policies may prevent schools from teaching and modeling principles of responsible use.
The surgeon general’s advisory is a well-intentioned response to real concerns about children’s well-being. But just as we wouldn’t hand a teenager the keys to a car without first teaching them how to drive, we can’t expect young people to navigate powerful AI technologies without purposeful guidance from an experienced instructor.
Already behind the wheel
Students need to achieve a certain degree of AI literacy. This includes a foundational understanding of what AI tools are, how they work, where they fall short, and who is accountable when they get things wrong, before they can use AI tools safely and effectively.
The public education system is ideally positioned to deliver that training in the same way that many states opt to offer driver’s ed in high schools. The main difference is that, when it comes to AI, most students who access the internet are already behind the wheel.
Across the country, students are using AI all the time. Artificial intelligence powers the results they get from search engines, the tools they use to do their homework, and the apps they open every day.
The question is not whether students use AI, because they already do. The question is whether educators will be allowed to help them use it well.
Many students are asking themselves when and how to best use AI and thinking critically about the answers. They want to know whether to trust what AI tells them. They want to know who owns what AI creates. They want to know what data it was trained on and what it doesn’t know.
These are not abstract philosophical questions. They are the same critical thinking skills we ask students to apply when evaluating a news source, interpreting a graph, or assessing a primary-source document.
The real risk
In many ways, AI literacy is not really a new subject. It is an extension of the habits of mind educators have always tried to build.
Through discussions of everything from plagiarism, misinformation, and personal data to bias, ethics, and AI’s impact on the environment, AI literacy coursework can equip students with the skills necessary to use the technology confidently, safely, and well—and to use more intellectual discernment in all areas of their lives.
The real risk of students using the technology without guidance is that they might never acquire the skills they need to evaluate what the AI tools are telling them. When a student accepts an AI output uncritically by failing to fact-check, by missing the bias embedded in the training data, or by not recognizing the limits of what the tool actually knows, that is a literacy failure. And right now, we are producing those failures at scale.
Technology deployed without purpose, without quality infrastructure, and without skilled educators behind it does not produce better outcomes. The evidence on that is clear.
But ineffective edtech cannot be combated by less technology. It can only be fixed through better implementation and a more honest evaluation of what works.
Logic does not hold
At this critical moment, sound policy means supporting AI literacy as a cross-disciplinary priority, woven into science, social studies, English language arts, and civics. It should not be treated as a standalone add-on.
It means investing in educator training so that teachers can facilitate these conversations with confidence. It means developing quality standards for AI tools used in learning environments. And it means engaging families, who deserve honest guidance about what their children are encountering and what schools are doing to prepare them for it.
Instead, from our perch, we see states and districts pursuing strategies that contradict one another. Policymakers are pledging to close the gap between K12 learning and workforce readiness while simultaneously moving to restrict or ban the very tools students will be expected to use on their first day of work.
The logic does not hold. Treating all screen exposure or AI use as equivalent oversimplifies how and why students are using technology in the first place.
In a country where highways connect major cities, we cannot ask students not to drive. We can only teach them how to do it safely.
The surgeon general is right that children’s well-being demands urgent attention. But enforcing a contrived separation between students and the technology that they will spend their entire adult lives navigating is not protection. It is avoidance.
In a labor market that is restructuring around AI capabilities, this avoidance has a cost and it will fall heaviest on the students who can least afford it.
