The Classroom Leap

How Estonia’s AI Experiment Is Reshaping European Education

eyesoneurope

Amsterdam, July 15, 2026 – Thirty years ago, a small Baltic nation of 1.36 million people made a bet that computers and the internet belonged in every school, long before most of Europe took the idea seriously. That bet, known as Tiger Leap, turned Estonia into “e-Estonia” — one of the most digitally advanced societies on the planet. Now Estonia is making a similar wager, and the rest of Europe is watching closely.

The Leap That Started It All

The AI Leap’s scope is ambitious given Estonia’s population of 1.36 million: over two years, the country plans to train 48,000 students and 6,700 teachers. The programme launched on September 1, 2025, giving roughly 20,000 students in grades 10 and 11, alongside thousands of teachers, structured access to advanced AI tools for the first time. By September 2026, it expands further, reaching vocational schools and a new cohort of 10th-graders.

What makes AI Leap notable isn’t just its scale — it’s the reasoning behind it. Organizers point out that 64 to 90% of Estonian students were already using AI tools before the programme even began, mostly free commercial apps used to speed through homework. Rather than fight that reality, Estonia chose to shape it. As the country’s president put it, the goal isn’t to use AI as much as possible, but to learn to use it wisely.

The tool built for Estonian classrooms, developed with OpenAI, deliberately doesn’t hand students ready-made answers. It’s designed to behave more like a Socratic tutor — asking guiding questions, nudging students toward their own reasoning, and responding only in Estonian outside of language classes. Early numbers suggest steady, organic uptake rather than forced adoption: by March 2026, thousands of students had activated accounts, with roughly half using the app weekly.

Crucially, the programme is layered, not just technological. It combines teacher training circles that meet monthly, a shared resource platform, regional coordinators spread across the country, and ongoing research led by the University of Tartu to track what actually helps students learn versus what just feels efficient. It’s this combination — technology, pedagogy, and evaluation moving together — that has other European governments taking notes.

A Continent Starts to Follow

Estonia’s experiment has become something of a reference model. A recent European Educational AI Index, which mapped strategies across Spain, Italy, France, Austria, Germany, and the UK, singled out Estonia as one of the continent’s clearest cases of technological integration and innovation, alongside honorable mentions for Poland and the Netherlands.

Poland has since rolled out its own national AI Development Policy, funding thousands of AI-equipped classrooms and pilot labs. France has introduced mandatory short AI-literacy sessions for secondary students through the Pix platform, alongside clearer rules on when and how students may use AI autonomously in class. The details differ from country to country, but the underlying instinct is the same one Estonia acted on first: don’t wait for AI to arrive in classrooms uninvited — build the guardrails and the pedagogy before it does.

The Netherlands: A Quieter, Research-Driven Path

The Netherlands hasn’t launched a single flagship programme the way Estonia has. Instead, Dutch AI-in-education efforts have grown through research consortia, university partnerships, and targeted classroom pilots — a more decentralized, evidence-first approach that reflects the country’s academic strengths and its cautious, GDPR-conscious institutional culture.

The centerpiece is NOLAI (National Education Lab AI), coordinated by Radboud University and funded through the Dutch National Growth Fund. Active since 2022, NOLAI brings together schools, researchers, EdTech companies, and policymakers to co-develop AI tools that hold up to ethical and pedagogical scrutiny, backed by its own dedicated ethics board.

A few concrete examples of what that looks like in practice:

• Self-regulated writing support. Researchers at Radboud’s Behavioural Science Institute are piloting a system that tracks how students actually work through essay-writing tasks — every click, pause, and revision — and translates that raw activity into meaningful patterns like “planning” or “orientation.” Students see their own process reflected back on a dashboard, turning an invisible habit into something they can reflect on and improve.
• Semi-automated scoring systems, designed to lighten the grading load on teachers while keeping a human firmly in the loop on final judgments.
• Progress-tracking and task recommendation tools, piloted in a small number of schools, that let teachers follow the learning trajectories of an entire class at once — something Joep van der Graaf, one of the researchers involved, describes as fundamentally new: the ability to meaningfully track thirty students’ learning processes simultaneously, not just their final grades.

Beyond NOLAI, momentum is building on the policy and industry side too. The Netherlands has positioned itself as an unusually strong AI hub for its size, and Vrije Universiteit Amsterdam is hosting a dedicated “AI for Education” conference in late October 2026, explicitly built around bridging research and real classroom practice — including a global student contest on agentic AI in education.

Same Goal, Different Roads

What ties Estonia and the Netherlands together isn’t identical policy — it’s a shared premise: AI is already in the classroom whether governments plan for it or not, so the real choice isn’t whether to allow it, but how deliberately to shape it. Estonia went big and fast, betting on national scale and speed. The Netherlands has gone careful and layered, betting on research rigor and teacher trust.

Both approaches are, in their own way, a repeat of a lesson Europe has learned before with earlier waves of classroom technology: the tool matters far less than the thinking that goes into how — and why — it’s used.