Nearly every decade, a new technology arrives in schools with the promise of transformation. Each time, educators are required to adapt or risk not equipping students with the tools needed to be successful in a rapidly changing, global workplace. We saw it with the internet, laptops, and tablets, technologies that were brought into classrooms before the critical questions were asked. We are living through that moment again, only faster and with far higher stakes.
Why are we integrating artificial intelligence (AI) into K–12 education at such a rapid pace? Who benefits when student data are harvested to train generative AI? How do we ensure that AI enters our classrooms as a supplement to human engagement, not a substitute for it? And, most urgently, are we trading away the very skills that make education worthwhile—the capacity to analyze information critically, to exercise sound judgment, to adapt to the unexpected, and to communicate ideas with clarity and conviction—in exchange for the appearance of efficiency? These are the questions that every caregiver, educator, school board member, and policymaker should be demanding answers to before another policy is instituted.
The rush to embed AI into our children’s classrooms is accelerating at a pace that has outrun our wisdom. Superintendents are signing contracts with tech vendors. And students are sitting in front of tools that are actively learning from them, while school leaders are still figuring out how to protect students’ personal information. The disconnect between adoption and oversight hasn’t gone unnoticed. At least one state’s leaders are already sketching out what comes next. The goal, the authors of a Reference Guide for AI Integration in Education in the Commonwealth of Virginia say, is for AI to help cultivate critical thinking, ethical judgment, and iterative refinement.
Picture a student who submits a perfectly formatted, step-by-step solution to a multipart algebra problem. Every line is correct. The answer matches the key. By every traditional measure, this student has demonstrated success in math. Now picture that same student having copied the problem into a generative AI app, received that solution in seconds, and transcribed it onto the page. This is the gap teachers must be equipped to close, the gap between a student communicating understanding and completing a task. Students need the judgment to recognize when a tool genuinely serves their learning, the discipline to use it thoughtfully, and an understanding of the mechanics, so they can build on what the tool provides. What we cannot afford is a generation that clicks “generate” without any sense of what’s happening beneath the surface.
Teachers often can’t tell when students have used AI, and without formal guidance, they’re largely on their own in figuring out how to respond. Also, there’s little guidance for teachers on using AI for instructional support. Nearly 60 percent say they’ve received no direction on using AI to grade work or give student feedback, and just under half report the same when it comes to creating assignments, building instructional materials, or differentiating for students’ needs.
The question is no longer whether AI belongs in the classroom. The deeper question is whether our assessments still measure anything real, and whether they’re designed to encourage creative thinking.
The gap between what we test and what students need
For generations, mathematics assessment has treated a sequence of steps as proof of understanding. Apply the approved algorithm, arrive at the right answer, pass the test. That model was orderly and gradable. It is also increasingly obsolete.
Research with high school students found that while AI readily supplies step-by-step solutions, only students who critiqued and adapted those solutions demonstrated genuine problem-solving ability. Procedural fluency alone no longer signals the pinnacle of math achievement. The question is not whether students can compete with machines. It’s whether schools are measuring students’ ability to do the things machines can’t do: engage in a community of learners using inquiry-based learning.
What distinguishes student thinking from machine learning? The National Academies of Sciences, Engineering, and Medicine’s 2026 report Data and Computing in K–12 Education makes a direct case for core competencies that machines cannot replicate:
These are precisely the capacities students need when nearly every significant life decision—college, career, health, financial—is touched by data someone else has analyzed.
The stakes are real and specific. Students who graduate confident in procedures they didn’t develop enter higher education and the workforce with limited durable skills needed to effectively navigate a data-saturated world. Students may struggle to evaluate a misleading graph, question a flawed risk model in a loan offer, or make sense of public health data.
What school leaders can do now
Math educators are the adults most positioned to notice when a student’s work reflects no real thinking and most trusted to design experiences that demand genuine reasoning instead. Administrators and superintendents set the professional learning conditions, such as time, safety, and resources.
School leaders need not wait. They can begin redefining how student learning is recognized and rewarded. They can invest in supporting teachers to design, facilitate, and score insight-centered tasks—providing time and opportunities to integrate AI into assignments in ways that reveal student thinking and make AI use transparent instead of forbidden.
Knowing when to trust a result, when to question it, and how to explain your reasoning to someone else is what an AI-saturated world actually demands. It is also what a generation of students deserves to be taught. Showing your work has to mean making your thinking visible.
For more insights on the role of math in ensuring educational equity, subscribe to Just Equations’ newsletter.
