Schools spent years working to put a device in every student's hands. Now a growing number of leaders are asking whether students are on those devices too much — during the school day and for homework alike. The phone debate may have started the conversation, but it hasn't stayed there. The unease has moved to the laptops and tablets schools deployed themselves, and to the screen time that comes with digital assignments.
States are starting to act on it. Utah, Missouri, Alabama, and Tennessee are among those exploring policies on the safety and effectiveness of educational technology. The question is shifting from how much technology schools can get to how, and whether, it should be used during instructional time.
If we're going to measure something, though, minutes on a device is the wrong number. The better measure is whether the tools and methods students use — digital or not — are backed by evidence, and whether they're improving outcomes.
What we should measure instead
Screen time is easy to measure, easy to compare, and easy to argue about. But skepticism toward classroom technology is fairly new. For most of my years in state education leadership, we were working to get more technology into schools. Research on student equity and outcomes showed that at-risk students made measurable progress in one-to-one laptop programs, so the goal was one device for every student. The pandemic pushed districts there almost overnight as learning moved home.
Then came the rush for content. In a single year, as classrooms went remote, the number of digital tools the average district used jumped from about 900 to more than 2,200. Some of it worked well. A lot of it was never built to teach.
So it's understandable to ask how many hours a student spends on a device. But the better question is whether the tool has evidence that it improves learning, and how strong that evidence is.
The Every Student Succeeds Act already gives us a shared way to answer. ESSA sorts evidence into four levels, from strong studies showing a tool actually caused better outcomes down to a clear, research-based rationale for why it should. The system isn't perfect, but it's widely understood, and it gives a board something steadier to weigh than gut feeling and brand recognition.
The 2026 EdTech Evidence Report applied that lens to the tools districts actually use. It found that about 40% of purpose-built edtech carries research evidence aligned to an ESSA tier. For the general consumer tools that show up in classrooms anyway, that figure drops to 2%. What matters is where the time goes. A tool built to teach, with evidence behind it, does more for a student than one that simply holds attention. The K–12 leaders who understand that distinction are already a step ahead on the argument forming right behind it: the one about AI.
AI is where minute-cutting backfires
One of the clearest responsibilities they have right now is preparing students for workplaces being rebuilt around AI, and that readiness doesn't come from keeping students away from the tools — it comes from guided, supervised practice with them. This is where cutting minutes for its own sake starts to work against schools. A blunt push to reduce device time collides head-on with the skills these students will be expected to have.
The way through isn't more minutes or fewer minutes. It's measuring the right thing: whether students are learning to use AI safely and with judgment, where it genuinely helps their work, where it gets misused, and how to tell the difference. This is supervised, responsible use, not a free-for-all, with clear guardrails around student privacy and age-appropriate practice. It takes clear instructional goals and teachers who know how to bring AI into their practice. And because AI is moving into classrooms faster than the research that would tell us what works, it's judgment we have to teach deliberately, not something we can wait on.
That makes discipline more important, not less. Before adopting an AI tool, ask what you'd ask of anything: what claims can it actually support, how does it protect student data, and will it work with the systems you already run? Districts that build this habit on today's tools won't start from zero when the harder AI questions arrive. And they will arrive.
What we'll work through at ISTELive 26
Whether learning happens digitally, on paper, or with AI, the real question is whether the strategy is evidence-based and improves student outcomes.
That's what I'll dig into at ISTELive 26 in Orlando on Tuesday, June 30, from 10 – 11 a.m. We'll work through the evaluation framework from the Evidence Report, the one districts and states can actually apply, and turn it into plain talking points you can use the week you get home: how to read an evidence claim, what to ask a vendor before you sign, and how to make the case for a well-researched tool.
If you want the underlying data first, the full report is worth reading before you come. And if you'd rather just talk it through, the Instructure team will be at the booth with updates across the ecosystem, or you can join us offsite on June 28 for the Cuba Libre reception, where these conversations tend to get more honest anyway.
The point is to walk into your next board meeting or vendor pitch with better questions and a way to weigh the answers.
