Norway Moves to Ban Generative AI in Elementary Schools

Norway tightens school rules on generative AI

Norway is moving to largely ban generative AI tools in elementary schools, taking one of Europe’s clearest policy positions yet on how artificial intelligence should be handled in early education. The new rules are set to take effect at the start of the school year in late August 2026.

Under the policy described by The Decoder, students in grades 1 through 7, roughly ages 6 to 13, generally will not be allowed to use AI tools at all. In lower secondary school, covering ages 14 to 16, AI use will be permitted only cautiously and under supervision. Older students will instead be taught how to use AI appropriately.

The government’s reasoning is direct: core learning skills come first. Prime Minister Jonas Gahr Stoere said the most important task of school is to ensure that children learn to read, write and do math, and argued that uncritical AI use can cause students to skip important learning steps.

A broader push to restore fundamentals

The AI restrictions are not an isolated measure. They are part of a wider Norwegian effort to rebalance schools away from heavy dependence on screens and digital systems. Stoere linked the new rules to a decline in learning outcomes since around 2015 and said smartphones, screens and algorithms are among the factors contributing to the problem.

That framing matters because it positions generative AI not only as a classroom technology question but as part of a broader educational and social policy debate. Norway is not simply asking whether AI can help students complete tasks more efficiently. It is asking whether early and poorly structured use of these systems undermines the slower foundational work that schools are meant to enforce.

The government also plans to pursue legislation requiring municipalities to provide physical teaching materials in schools, effectively bringing more books back into classrooms. Stoere said previous governments had given too much weight to digital media. Norway had already banned smartphones in schools, increased teachers’ authority in the classroom and is planning a social media ban for children under 16.

Taken together, those moves amount to a deliberate rollback of the assumption that more digital access automatically improves education. In Norway’s current view, some digital tools may be useful later, but early schooling should focus more heavily on direct instruction, printed materials and core skill-building.

Why generative AI is getting special scrutiny

Generative AI has moved unusually quickly from novelty to everyday tool, and education systems are still deciding whether to treat it as a calculator-like aid, a research assistant, a plagiarism risk or some combination of all three. Norway’s approach suggests that for younger children, the downside risks outweigh the benefits.

The concern is not merely that students might cheat. It is that AI can short-circuit the developmental process itself. If children rely on generated text, summaries or problem-solving assistance too early, they may practice less reading, less writing and less independent reasoning. For a government already worried about declining outcomes, that is a strong enough case for restriction.

Norway is still leaving some room for AI in older grades, which shows the policy is not a wholesale rejection of the technology. Instead, it draws a developmental line. Younger students are to be shielded from use; middle-grade students can use AI only carefully and with oversight; older students should learn proper use rather than being left to discover it informally.

That tiered model may prove influential because it recognizes that the educational value and risk of AI change with age. The same tool that may weaken early literacy habits could later become a subject of digital literacy instruction in its own right.

Part of an international debate

Norway is not alone in tightening rules, though countries remain far from consensus. The Decoder cited several examples showing how fragmented the global response still is.

Japan issued guidelines in 2023 calling for special caution with children under 13 and classifying AI-generated schoolwork as cheating. In the United States, a court ruled in 2024 that schools can penalize unauthorized AI use. UC Berkeley Law School is set to ban AI for nearly all graded assignments starting in the summer of 2026, allowing it only for research.

At the same time, other governments are moving in the opposite direction. The United Arab Emirates will make AI a required subject from kindergarten through 12th grade beginning in the 2025-26 school year. In Germany, the Conference of Ministers of Education has argued for integrating AI into the classroom and called an outright ban unrealistic.

Those differences reveal a deeper divide in education policy. One side sees AI primarily as an inevitable capability that students must learn to use early. The other sees it as a powerful tool that needs to be delayed, limited or tightly structured until basic competencies are secure. Norway has now placed itself firmly in the second camp.

What the decision could mean

Norway’s move will likely be watched closely across Europe because it offers a concrete regulatory model rather than broad guidance. The timing is also notable. Many school systems have spent the past two years issuing provisional advice while teachers and administrators adapt in real time. Norway is going further by establishing age-based rules backed by a clear political argument about learning quality.

Whether the restrictions improve outcomes will depend on implementation. Schools will need workable definitions of what counts as an AI tool, practical supervision rules for older students and classroom policies that teachers can enforce without excessive administrative burden. Still, the intent is unmistakable: AI should not displace the foundational work of learning.

For the wider AI policy debate, the Norwegian decision adds momentum to the idea that regulation may vary sharply by age and context rather than by technology alone. In other words, the question is becoming less about whether generative AI belongs in education at all and more about when, where and under what conditions it should be introduced.

As the late August rollout approaches, Norway will become an important test case for a school system trying to put limits around AI before those tools become fully normalized in everyday classroom practice.

This article is based on reporting by The Decoder. Read the original article.