Imagine the Organisation for Economic Co-operation and Development (OECD) publishing a report tomorrow showing that the GDP of virtually every country on earth—rich and poor, east and west—has been falling year upon year for a decade. Imagine further that it uses a single headline for this report: “Future-Ready Economies.”
Imagine that it does not pause to ask, with any seriousness, why the wealth of nations is shrinking in near lockstep across wildly different political systems and policy regimes, offering instead a buffet of generalized explanations—demographic, institutional, a touch of pandemic residue—general enough to explain everything in the world and therefore nothing at all.
And on the one force that has entered every country over this exact period, in roughly the same form, it opts to equivocate, murmuring that its effects “depend on how it is used.”
We would call this what it is: a professional and moral failure. An institution with unmatched data and high credibility choosing not to say in the clearest language that something has gone horribly wrong.
The equivalent of this is what the OECD has just done with PISA 2025 results—the largest global survey to assess the health of school education by testing 15-year-olds.
Consider the numbers. Across OECD countries, reading scores have fallen by 28 points since 2015—which is well over a year’s worth of learning. Mathematics is down 22 points over the same period. Reading has fallen from a peak of 501 in 2012 to 466 last year; mathematics from 502 to 469; science is relatively resilient, with a fall from a 2009 high of 506 to 486.
These are the lowest scores recorded since PISA began in 2000, with the latest survey covering 91 countries and 33 million 15-year-olds.
By any standard, this is clearly a global emergency in the education of the young.
Yet, look at what the OECD chose to call it. Not “Global School Education in Unprecedented Crisis.” Not even “A Warning to the World.” The report is titled, astonishingly, “Future-Ready Students.” This is the euphemism of an organization that has the data to sound an urgent alarm but instead reaches for anodyne obfuscation.
The report’s own explanations compound the offence. It points to teacher shortages, pandemic disruption, the general fraying of school systems and falling rates of reading for pleasure. Each may carry a grain of truth in some country. But notice what none of them can do individually or combined: explain a decline that is secular and near-universal, cutting across countries with utterly different curricula, teacher-training regimes and levels of public investment.
Nothing uniform has happened to curriculum design in Finland and Costa Rica and France in this period. Teacher shortages are not identically distributed across 91 education systems. Pandemic disruption cannot explain declines that, as the report itself concedes, predate 2020 in many countries.
When an explanation is called upon to account for something happening everywhere all at once, and that too in systems sharing almost nothing else in common, and the report offers little beyond a dozen locally-flavoured hypotheses, it cannot be called an analysis. It’s a side-stepping dressed up as nuance.
There does exist a single thing that has entered every child’s life in each of these countries over the past decade, and in broadly the same form: the smartphone, and with it the algorithmic feed and now the AI chatbot—inside the school gate and outside it.
The report’s own data gives this away even as its prose tiptoes around it. More than a quarter of students say classmates are distracted by devices in most or every science lesson, and those surrounded by distraction perform worse and feel less connected to school.
Nearly half use AI chatbots regularly for schoolwork; those who use them to draft essays and summarize texts score roughly 20 points lower in science than those who do not.
Reading for pleasure—the deep, unhurried, attention-building kind—is falling wherever screens are proliferating, and the countries with the steepest reading declines also post the steepest falls in mathematics and science, exactly as one would expect if the loss of sustained, effortful reading has begun to corrode the cognitive muscles every other subject depends on.
This is not a minor footnote. It is the closest thing to a smoking gun in a social-science survey of this scale. And still, the report insists on straddling the fence. Technology and AI, it says, “can strengthen learning... but only when used purposefully,” a sentence so hedged that it could sit comfortably in a phone manufacturer’s press release.
When the correlation between AI use and weaker performance is so consistent, so large and so global, such equivocation is not scientific caution but abdication.
The OECD is not a marginal actor here. PISA is the most credible, longest-running, most methodologically robust cross-national measure of what children learn. With that authority comes an obligation that goes beyond publishing tables and averages: to name a crisis when the evidence points overwhelmingly to one.
Governments will not be moved by hedged words and parents won’t restrict devices on their basis. What this moment demands is a global alarm raised, one that is decisive, urgent and unafraid to say plainly that we are failing an entire generation in school after school and country after country—even as the clock runs on the scope for preventive action.
The author is CEO of Azim Premji Foundation.