Artificial intelligence is no longer a distant promise in education - it is already inside the classroom, on the laptop, in the homework. A 2026 OECD-cited survey of more than 7,000 12-to-17-year-olds across seven European countries (Germany, Greece, Portugal, Romania, Spain, Türkiye and the UK) found high use of generative AI among students, with the most common uses being looking up information and getting explanations of concepts. Among teachers, the OECD's own TALIS data shows that 37% of lower-secondary educators reported using AI in their work in 2024 - a number that has almost certainly grown since. This is not a future trend. It is a present-tense transformation, and it is moving faster than most school systems can formally respond to it.

The opportunity is substantial - but it is not unconditional. The OECD's Digital Education Outlook 2026 makes a point that should temper any breathless optimism: generative AI improves learning only when it is guided by clear pedagogical principles. Used without that guidance, it can simply make students look more productive - better-written essays, faster homework - without producing any real learning gain. In other words: AI does not automatically make people smarter. It makes the gap between guided and unguided use enormous.

Used well, though, the upside is real. Intelligent tutoring systems can deliver personalized support even in low-infrastructure settings, closing - rather than widening - access gaps. Teachers, meanwhile, can offload repetitive preparation work and redirect that time toward mentoring, feedback and the kind of judgment a model still cannot replicate. This is not about replacing educators. It is about giving them back their time.

For Employers, the Shift Is Already Showing Up in Salary Data

For employers, the shift is already showing up in salary data, not just survey sentiment. A 2026 World Economic Forum analysis of more than 10 million UK job postings found that candidates with AI-related skills command an advertised salary 23% higher than otherwise comparable candidates without them - a premium that now outpaces the wage bump from a Bachelor's degree (8%) and rivals that of a Master's (13%). AI skills, in other words, have started to out-earn formal credentials. And PwC's 2026 Global AI Jobs Barometer, built on more than a billion job ads across six continents, found that companies most exposed to AI are growing productivity 40% faster than those least exposed - with the top 20% of AI-adopting firms posting productivity growth of 163% since 2018, nearly five times the average.

There's a second, less comfortable number underneath this: the Future of Jobs Report 2025 found that 41% of employers expect to cut roles vulnerable to AI-driven obsolescence, while 70% simultaneously plan to hire for new AI-related skills. That contradiction - shrinking and hiring at once - is the labour market's way of saying the kind of human contribution is what's changing, not the need for humans.

Global Institutions Agree on Where That Contribution Lies

Global institutions agree on where that contribution lies. The OECD, UNESCO, Stanford and the WEF converge on the same shortlist of durable skills: critical thinking, collaboration, creativity, self-regulated learning, communication. None of these are new ideas - what's new is the urgency. WEF's 2025 employer survey found that analytical thinking is now the single most in-demand skill, cited as essential by seven in ten companies, ahead of resilience, leadership or any specific technical competency. The strongest advantage will go to people who pair human judgment with AI fluency - not to those who outsource one for the other.

That is why education has to shift from passive knowledge transfer to active skill formation. Knowledge itself is no longer the scarce resource - it is one prompt away. What's scarce is the judgment to use it well: knowing when an AI answer is right, when it's confidently wrong, and when the real learning happens in the friction the AI just removed. Used well, AI can build research habits, sharpen communication, and create the kind of lifelong-learning discipline that no classroom lecture ever could on its own.

There's a striking parallel worth sitting with: economists studying AI's labour effects keep reaching for the same historical comparison - electricity, the automobile, the spreadsheet. Each technology did not just speed up old tasks; it reorganized which skills mattered at all. The spreadsheet did not eliminate accountants - it eliminated manual ledger-keeping and created a generation of analysts. AI in education may be following the same arc: not erasing the teacher or the student's effort, but relocating where the effort goes.

The future of schooling will not be defined by machines replacing people. It will be defined by the schools that taught their students to think with the machine rather than through it - and by the ones that did not.

For the labour market, this is a defining moment, and the data already shows the shape of it: a 23% wage premium here, a 40% productivity gap there, and - by 2030 - a projected 170 million new jobs created against 92 million displaced, for a net gain of 78 million roles globally, according to the WEF's Future of Jobs Report 2025. AI in education is not just a classroom innovation. It is the foundation the next decade of careers will be built on - for those who get there early enough to build on it.

Sources

OECD Digital Education Outlook 2026; OECD TALIS 2024; World Economic Forum, "How AI skills and experience are transforming the workplace" (2026); PwC 2026 Global AI Jobs Barometer; WEF Future of Jobs Report 2025 (170M created / 92M displaced / 78M net new jobs by 2030).