A Warning About My Own Track Record
I should be upfront. I have been building and investing in edtech businesses for over 25 years, so I have a reasonably shrewd idea of what works and what doesn’t. In that time I have probably read over 2,000 pitch decks and invested in fewer than 20 – a No-to-Yes ratio of 99:1. Of those, almost half went bust or screwed their angel investors and moved on.
I could easily claim that what works in edtech,from a business perspective as distinct from a pedagogical one, conforms to my POPEYE model, but that isn’t true. I can name 20+ companies who ‘successfully’ sell to schools, but who are neither profitable, built on solid pedagogy nor supported by robust evidence. So yes, I have skin in this game and have plenty of biases (which I’m always working on). Bear that in mind. But it also means I have watched the failures and success in edtech up close, and the current anti-edtech narrative misreads almost all of them.
Three Things Being Treated as One
The current discourse conflates three distinct issues
- Disruption caused by mobile phones in schools
- The broader social issues around the use of social media and screens outside school
- The efficacy and opportunity cost of edtech in education from national systems down to classroom and student level.
These are not the same problem, and treating them as one produces bad policy.
On mobile phones in schools, the campaigners have a point. Devices that can’t be tightly locked down are distraction engines and growing prohibitions on mobiles in schools are a necessary step, as much about managing the learning environment (especially disruptive behaviour) as reducing techno distraction.
The workplace is moving in the same direction: JP Morgan Chase’s Jamie Dimon has called mobile use at work disrespectful and time-wasting and I would apply similar restrictions to civil servants at every level. However, some workplace researchers argue that the evidence for blanket phone bans is weaker than proponents claim, and that employees perceive them negatively when mobile access has obvious professional utility. An issue worth watching.
On social media, the campaigns have identified a real problem but proposed a poor solution. Australia’s ban on social media for under-16s has been instructive. I have yet to speak to a parent who believes it has prevented their child from being on social media, and the kids say the guardrails are easy to circumvent. A VPN costs $1.78 per month from SurfShark, and Firefox now has an inbuilt VPN. This rather undermines the claims of Australia’s ESafety Commissioner, Julie Inman-Grant, that VPNs aren’t impacting access to social media because, “the costs are in the thousands of dollars”. The policy puts penalties on platforms rather than families, politically convenient but structurally inadequate. The UK and other countries will soon replicate it with similarly tenuous results.
Edtech is the software, platforms and systems used in teaching and learning. The growing number of attacks are frequently made without credible evidence, and their popular narratives are riddled with errors, omissions and biased interpretations of what they define as ‘evidence’.
The Nordic Data Is Far More Complicated Than Claimed
The anti-edtech poster child is Norway, which had a significant national edtech rollout in its schools from the mid-2000s. Norway has recently reversed course sharply. From August 2026, Norway’s new national regulations will require schools to be “particularly careful” with screen use for all children in the first four years of school. The government is even considering removing “digital skills” from the five core pillars of the national curriculum – the others are reading, writing, maths and oral skills – and replacing it with a standalone technology subject introduced at the secondary level. Prime Minister Jonas Gahr Støre has stated that 15,000 pupils currently finish primary school unable to read properly. Former education minister Trine Skei Grande, now Director of the Norwegian Publishers’ Association, has put it bluntly: “We are far, far too rich, so we do stupid things with our money”.
But Sweden is the more instructive case study. Sweden introduced edtech into its pre-school curriculum officially in 2019, then terminated that policy in 2025. It has moved to a “from screen to binder” approach backed by £157m in grants (approximately £142 per student but over several years) and a revised textbook-focused curriculum from 2028. The Liberal party spokesperson, Joar Forsell, has said the policy is explicitly to “get rid of screens as much as possible.” The primary policy driver has been Sweden’s falling PISA results.
But there is a fundamental problem with that argument. Sweden’s significant PISA decline, particularly in 2012 and between 2018 and 2022 saw Sweden performing better than all of its Nordic contemporaries over the same period. What PISA 2022 reveals is not a Swedish edtech failure, but a regional systemic collapse. The Nordic model built on highly-trained teachers, no ability grouping (debunked by new EEF research) and minimal standardised testing. Yet their PISA rating has still fallen significantly. To lay the blame on edtech is not analysis; it is scapegoating and scaremongering.
Finland is the deeper story here. For two decades, politicians, advisors and educational tourists made the pilgrimage to Helsinki to understand why Finland’s system was “so superior” (something I call the “Pasi Sahlberg effect”). This was largely a false success narrative, comprehensively unpicked by Tim Oates in his 2015 Cambridge Assessment report Finnish Fairy Stories. Oates showed that foreign analysts had mistaken a cross-sectional PISA snapshot for longitudinal improvement, and that Finland in 2000 (the point at which the educational tourism began) was already on a downward trajectory. The “Finland has no national testing” mythology was precisely that: the Finland that produced those early PISA scores had highly centralised inspection and testing. The myth persists regardless, and it continues to underpin a large amount of anti-edtech thinking.
What the High Performers Actually Have in Common
The countries that do well in PISA, like Estonia and Singapore, share something the Nordic countries do not: a deeply centralised, system-wide approach to edtech embedded in teacher training and curriculum from the ground up, rather than devolving edtech to schools and districts to figure out largely on their own. The lesson is not less technology, but better embedded pedagogically-based technology, integrated by design rather than bolted on by procurement and adhoc ‘Professional Development’.
What “Edtech” Actually Is
Before any of this can be properly debated, it helps to define what we are actually talking about. In any school system today, 80% or more of the tech used has nothing to do with classrooms; it’s the infrastructure that runs HR, finance, timetabling, compliance, cybersecurity, legal, communications, energy, payroll, data, exams, transport, etc. The interesting and contested question is about the remaining 20%, and even within that the landscape is not what the campaigners suggest.
The device question is largely a distraction. Giving students an iPad, Chromebook or any device does not equate to educational improvement, because it conflates hardware with software, content, user behaviour and educator involvement – the latter three being what really matters. It doesn’t matter what device a student has if the edtech and content is poor, and in my experience most of it is. One much-vaunted app, Easy Peasy, had extraordinary initial supporting evidence until you read it carefully and found the research was biased and failed the basic opportunity cost test: was it better than the best alternative schools or parents already had? I have yet to see that question asked seriously in the current debate about edtech and especially about whether analogue tools are any better (most by their very nature don’t generate much data to evaluate).
On hardware specifically, schools that boast about their iPad programmes should examine their Total Cost of Ownership figures. The TCO of a Chromebook over its four-to-five year life is approximately three times lower than an iPad equivalent. High TCO means less money available for what actually determines outcomes: high quality content and deeper professional development for teachers. A recent anti-edtech story from the US was about a mother finding her son had watched 13,000 YouTube videos on his ChromeBook at school. That’s terrible but it’s a simple fix that shows the school/district needs a serious bollocking about not managing the basic cyber security on their network and devices (it’s arguably also a serious failure of pastoral care).
Three Issues the Anti-Edtech Movement Systematically Ignores
Teacher quality and professional development. Most education systems do a terrible job of CPD when it comes to edtech. Part of this is structural. In the UK and Australia (the markets I know best) I have despaired for 15 years at the failure of centralised and decentralised edtech procurement with little or no coordinated, pedagogically-based CPD strategy. Edtech companies share part of the blame. Their business models, in a competitive price-sensitive market, often leave little margin for proper onboarding and follow-up training. Rather than addressing this, they often claim that teachers are simply” not interested”. This masks a big and unpopular topic of teacher quality problem and the problems of performance managing out weak teachers who can’t be upskilled. The culture of “passing the lemon”, the chronic shortage of subject specialists in maths and languages and the international recruitment and retention crisis are far deeper systemic problems than the amount of edtech in classrooms, and the anti-edtech campaign are silent about all of it.
Assessment and data. Moving away from edtech means moving back towards either large amounts of formative teacher assessment (a known area of significant inaccuracy and a substantial workload burden) or high-stakes summative testing through the likes of NAPLAN, Key Stage tests, PISA, TIMSS and their equivalents. These are expensive and, with some irony, increasingly conducted online themselves (large-scale, invigilated, in-school tests are expensive). Edtech can provide swift and granular formative and summative assessment data at individual, class, school and sometimes national level.
Dashboards from individual edtech providers have been hard to coordinate and some of the data they collect is pointless and can represent a safeguarding problem. This weakness has improved markedly over the last four years, but the fundamental challenge I see is that less edtech means more teacher workload and stress. This is hardly a solution to the recruitment and retention crisis, likely to have a more negative impact on student outcomes than any amount of edtech.
There is also a glaring absence in the literature, with no serious analysis of the Opportunity Cost of replacing edtech with analogue resources. Take books for example; every educational publisher I know privately acknowledges that their books and schemes of work are almost never substantially revised between editions and that there is essentially no usage and outcomes research or evidence-base, of the kind demanded of edtech. The evidential standards applied to the two sides of this debate are not remotely equivalent.
The 80% of the year spent outside school. Students are in school for at most 20% of each year. The notion that prohibiting tech for that 20% will deliver substantive educational and social benefit is deluded if the other 80% is unaddressed. Parents purchase most of the devices and data plans used outside school, particularly mobile phones. Parental responsibility is systematically underplayed in the campaign discourse, which is politically understandable but intellectually indefensible.
A Genuine Success Story the Campaigns Ignore
One of the most significant edtech successes of the last decade is Oak National Academy, which began during Covid and became arguably the world’s most successful online learning platform during and post Covid. It is the only fully open source national curriculum resource of its kind, published under the Open Government Licence (compatible with Creative Commons), meaning any organisation can build on it freely. Around a third of all teachers in England now use Oak products. Its AI lesson assistant, Aila, launched in 2024, operates with the most rigorous safety and compliance architecture of any AI tool currently in the edtech market combining fully open source code, human-in-the-loop lesson planning, curriculum-anchored retrieval-augmented generation to reduce hallucination, and detailed algorithmic transparency documentation published under the UK government’s own standards. It was the first publicly-funded, generative AI tool made available to the public in the UK. This is what good, system-embedded, transparently-governed edtech looks like and it gets almost no attention from the campaign groups.
The Celebrity Campaign: the Dunning-Kruger Effect With a Media Budget
The two organisations driving most of the UK anti-edtech campaigns are Close Screens Open Minds (registered as Close Screens Open Minds Limited (CSOM), company number 16965274) and Safe Screens (operating through UsForThem Limited Company number 13337939).
CSOM deserves particular scrutiny, starting with its funding. The organisation’s website footer states that it is funded by “patron and founders Anna and Hugh Grant”, so this is a for-profit, single director , personally-financed campaign vehicle for Grant and his partner Anna Eberstein. Grant can fulminate about “pathetic rules” affecting his children all he likes, but his and co-patron Sophie Winkleman’s opprobrium would be better directed inward. Both send their children to expensive private schools, Grant has described the schools his children attend as “ridiculously expensive”, and those same schools all operate 1:1 device programmes! This is the luxury belief problem in its purest form; people wealthy and well-connected enough to absorb the consequences of their own stated positions.
The advisory and expert structures around CSOM warrant equal scrutiny. Their campaigns have attracted some credible names. But it has also assembled a roster that, examined carefully, raises real questions about the conflicts of interest involved. For example Julie and Andrew Liddell, co-founders of the EdTech Law Center, claimed to have, “addressed the UK Parliament in November 2025” as part of the CSOM operation whereas they actually spoke at a side event linked to the Education All Parliamentary Education Group. APPG’s are not parliament; this specific one is funded and effectively run by the British Educational Suppliers Association, whose members include Microsoft and many well-known edtech companies. The EdTech Law Center is, by their own description, a “consumer protection law firm” that helps families sue edtech companies albeit in the US and not the UK. They are not researchers or academics; they are activist lawyers with a direct commercial financial interest in the growth of anti-edtech litigation. Presenting them as independent experts on the evidence around edtech is misleading.
Another expert is Ane Vernon, a partner at Payne Hicks Beach in London, and she does bring very serious UK legal credibility to the campaign’s positioning. Payne Hicks Beach is one of the most exclusive private client law firms in England, operating from Lincoln’s Inn and serving high-net worth individuals, including, as Vernon’s own profile makes clear, ‘private school parents involved in education disputes’. Vernon has written pieces on “the vulnerabilities of school-issued digital devices” and “safeguarding in the digital classroom”, but she is not a neutral authority, she’s an education disputes lawyer whose client base is precisely the demographic that can afford to fund and sustain anti-edtech/social media/mobile device campaigns.
The broader problem here is absence of disclosure. A campaign positioning itself as an “evidence-based resource” should be transparent about their interests – commercial, legal and personal – and to my mind at least many do not. One of CSOMs experts is Professor Jonathan Haidt, someone genuinely worth engaging with on adolescent mental health and social media. But by attaching his scientific credibility to a celebrity media operation funded by wealthy private school parents, supported by litigators and others with financial stakes in the outcomes, he risks becoming more counter-productive than beneficial. His flagship work The Anxious Generation, which the campaign deploys heavily, has already been subjected to substantial academic criticism from Professor Candace Odgers in Nature, from Ben Chugg’s forensic analysis, and from Andrew Marcinek’s Substack. The conclusions are considerably more contested than the campaign’s confident deployment of them suggests.
The Underlying Question Nobody Is Asking
The debate about edtech efficacy matters, but it will not be settled by bans. Prohibition did not work for alcohol and it will not work here as Australia’s social media ban already demonstrably shows. Grant and Winkleman’s children will barely notice either way, because their parents’ wealth and social capital will offset whatever educational ground is supposedly lost. The children who will be affected are the ones in state schools, whose parents may not be able to compensate.
What the debate actually requires is honest analysis: rigorous opportunity cost comparisons between edtech and analogue alternatives; a serious look at the centralised, system-embedded models that work in Estonia, Singapore and in its own way Oak National Academy; and an honest reckoning with teacher quality, professional development and the 80% of children’s lives that schools do not control. Until the campaign groups are willing to engage with those questions, and until they are transparent about the interests of those asking them, what we are watching is not an evidence-based movement, it is a well-funded, celebrity-driven, media operation.
