Norway imposes near ban on AI in elementary school

Policy details & context

• Norway’s guidance is age-graded: 6–13 generally should not use AI; 14–16 may use it cautiously under supervision; 17–19 should learn appropriate AI use for work and higher education.
• This follows earlier moves: banning smartphones in schools and strengthening classroom discipline powers in response to falling test scores.
• Norway had heavily digitized classrooms over ~20 years and is now partially unwinding that, which some see as “following the evidence” on screens and learning.

Arguments supporting the near‑ban

• Elementary years are for building core skills: reading, writing, arithmetic, attention, frustration tolerance, and independent thinking. Generative AI is seen as undermining these, especially if used as a shortcut.
• Multiple commenters liken AI to a “forklift at the gym”: students move the weights (finish homework) but don’t build the muscles (understanding).
• Fears: reduced critical thinking, overreliance on AI, long‑term cognitive atrophy, and a self‑fulfilling “we need AI because we’ve forgotten how to think” scenario.
• Some argue it’s analogous to calculators: you don’t introduce them before kids can do arithmetic unaided.

Concerns and criticisms

• Several worry this is like 1990s internet bans or early calculator bans: over‑cautious and potentially regressive.
• Others fault the lack of public nuance: they want clearer messaging that this is temporary while research and guard‑railed tools improve.
• A minority argues traditional schooling feels threatened by AI “outcompeting” it.

Evidence and research referenced

• Multiple studies (Nature, CEPR, MIT, Microsoft) are cited claiming:
  • AI use often boosts homework scores and speed.
  • But exam scores and deeper comprehension decline over months to years, including for “careful” users.
• Commenters repeatedly stress: good evidence of harm in many settings; no comparably strong evidence yet of net benefits for learning, especially in children.
• Pro‑AI voices counter with personal anecdotes of faster learning, but are challenged for lacking external benchmarks.

How AI is currently used in schools

• One Norwegian parent describes 10–13‑year‑olds using ChatGPT for:
  • Brainstorming and “cold start” help on writing.
  • Getting feedback on drafts.
  • Generating entire speeches/presentations.
• Kids themselves reportedly admit they mainly use AI to avoid doing homework, not to understand it.
• Teachers increasingly:
  • Use AI to create lesson plans, worksheets, and assignments.
  • Sometimes produce incorrect material and assume AI is “foolproof.”
• Some describe a full loop: AI creates assignments, students solve with AI, teachers grade with AI, administrators monitor with AI—seen as an “education charade.”

Broader tech‑in‑education debate

• Several argue screens and 1:1 devices generally hurt attention and outcomes; some call for banning or sharply limiting computers in class, returning to pen‑and‑paper and in‑person exams.
• Others distinguish between:
  • Unrestricted general models (seen as harmful), and
  • Highly constrained, pedagogically designed tutors, which they believe could deliver 1:1, adaptive instruction—though data for this is not yet public.

Equity, parenting, and enforcement

• Wealthier families already tightly restrict device use; bans may mostly affect poorer kids who rely on school devices.
• One view: a school ban only removes the “equalizing” access; affluent kids will still learn AI at home.
• Counter‑view: given current evidence of harm, society shouldn’t “experiment on kids” to protect hypothetical future AI‑fluency advantages.
• Enforcement is seen as hard once students have phones and home access; many predict a shift back toward supervised, in‑class work and high‑stakes exams.

Meta‑discussion

• Some notice strong anti‑AI sentiment and downvoting of pro‑AI comments, while others frame the backlash as a data‑driven correction to hype rather than “Luddism.”