An associate who just returned from the recent Frankfurt Book Fair gushed about the exciting things happening in the edtech content space, and I was reminded of the pushback against technology in education we’re seeing, with countries like Sweden rolling back on digital devices and prominent campaigners like Sophie Winkleman forcefully arguing against the negative effects of technology on student learning. Malaysia is nowhere near taking back devices as teachers complain there aren’t enough in schools to begin with. Perhaps it’s an opportune time to take stock of technology’s role in education, as AI isn’t making it any simpler.
A middle school in central Kansas asked its 480 students to hand back their school-issued Chromebooks last December, and the laptops now sit in carts at the back of the classrooms while the children take notes on paper.
The school is McPherson Middle School, and the sequence of events, as Fortune reported it (drawing on the New York Times), is instructive. It had banned mobile phones four years earlier and left the laptops alone, and once the phones were gone the administrators found that students were using the laptops for what the phones had been used for: watching YouTube, playing games, and in some cases teasing other students through their school email accounts. Laptops are now used only for specific activities that a teacher assigns, and the principal, Inge Esping, put the position plainly: “This technology can be a tool. It is not the answer to education.”
It is not only one school in Kansas
According to the same Fortune report, schools in North Carolina, Virginia, Maryland and Michigan are also reconsidering policies that put a laptop in every student’s hands. The reasons vary, and it is worth keeping them apart. In North Carolina some of the pressure is financial: the state’s schools spent $448 million of pandemic-related federal funding on computers and equipment, the money has run out, and devices that last less than a decade are now due for replacement. The superintendent of Wake County, the state’s largest school system, said in 2025 that the district needed to move away from its one-to-one policy. In Burke County a school board resolution encouraged paper and printed materials and limited screens to activities where technology offers clear, evidence-based advantages, and parents and teachers there reported improved reading comprehension and test scores by February (the report relies on their accounts and does not give figures). In Wexford County, Michigan, a district banned screens for elementary pupils after more than 65 per cent of the third, fourth and fifth graders in one school were rated “not proficient” or “partially proficient” on state tests.
The evidence behind the reversal deserves some caution, and Fortune itself says the studies “reportedly” coincide with falling or flat scores. The neuroscientist Jared Cooney Horvath, in written testimony to the US Senate, cited the international TIMSS mathematics and science study to argue that frequent in-class computer use correlates with lower performance. A correlation of that kind does not show that the laptops caused the decline, and Maine’s experience, where scores did not improve after fifteen years of a statewide laptop programme, shows no gain, which is a different finding from showing harm. For a school leader, though, the practical point survives the caveat: the assumption that a device for every child would raise results has not been borne out where it has been tested for long enough.
Why this matters before the AI purchases begin
The argument for pausing is made with some force in a July opinion piece in The 74 by Liz Cohen of 50CAN, who compares a screen ban to Dry January: an imposed break that lets a school ask what it actually misses, what worked, and which products produce learning. She notes that Utah, Tennessee and Iowa had limits or bans on screens in the early primary grades as of July, and that when a ban lifts, technology can be brought back slowly, with clear measures of whether it helps. Her proposal is an opinion and not a finding, and she is writing about the American policy cycle, but the sequence she describes (stop, ask, reintroduce with a purpose) is one I would want any school to follow before adding AI tools to a device programme it has never properly reviewed.
I made a similar point in the piece on who should teach AI literacy: a school that cannot run a good lesson without the vendor’s platform has not built a capability, only a dependency. The laptop reversals are what that dependency looks like several years on, when the funding has gone and the results are unconvincing. AI tools arriving through the same devices, often as a free feature of a platform already in use, will inherit whatever habits the devices created.
Before adding any AI tool, ask your teachers for one list: the lessons in which a device clearly improved learning this term, and the lessons in which it was open but not needed. If the first list is short, the problem to solve is the programme and not the tool. Then check that every important lesson has a version that can run on paper if the platform fails, changes its price or is withdrawn.
Why this matters for Malaysian schools
I have not seen comparable reports of Malaysian schools returning their devices, and I would not suggest the American pattern will repeat here. The risk worth noticing is of a different kind. With the 2027 curriculum bringing new expectations for digital competence, there will be pressure to show visible adoption quickly, and visible adoption is easiest to buy: devices and licences can be counted, whereas judgment in a classroom cannot. The American schools did the counting first and the questioning later. Doing it the other way round costs very little.
Want an independent view before the next purchase?
The 90-minute SLT Briefing gives your leadership team a vendor-neutral look at where devices and AI belong in your programme, what can run unplugged, and what the 2027 mandate actually asks of you.
Book a Principal's BriefingWhat This Means for You
A device in every pair of hands was once treated as proof of a school’s seriousness, and the schools now reversing course are not rejecting technology so much as discovering that nobody had asked what each device was for. Ask the question now, while the AI purchases are still ahead of you, and the answer will probably be a smaller, better-chosen set of tools than the market would like you to buy.