Prime Minister Jonas Gahr Stoere's government will bar generative AI tools for students aged 6 to 13 starting late August, with supervised use for 14 to 16 year olds and encouraged independent use from 17 up.
Norway is about to run the same policy experiment twice, this time on a technology that did not exist when the first one started.
The Norwegian government will bar students aged 6 to 13 from using generative AI tools in school at the start of the 2026-2027 school year in late August, Prime Minister Jonas Gahr Stoere said at a press conference. Students aged 14 to 16 will be allowed to use generative AI only under teacher supervision. Those aged 17 and older are encouraged to use AI "appropriately" on their own. The restrictions are Norway's first major attempt to govern classroom AI use, and they are an explicit extension of a policy design that has already been tested on a different device.
That earlier test was the 2024 ban on smartphones and tablets in Norwegian classrooms. In the year since, the policy has been credited, in wire reporting and by Norwegian officials, with measurable improvements: reduced bullying, better grades, a significant drop in screen time, and a decline in psychologist visits, with the largest gains among girls. The government is now treating those outcomes as evidence that an age-graded, technology-specific classroom restriction can work, and is applying the same template to a new set of tools.
Stoere's stated rationale, that AI lets children "skip crucial steps" in learning and that schools should prioritize teaching them "to read, write and do mathematics," echoes the framing used when Norway rolled out its phone ban two years ago. The mechanics are deliberately familiar. The new policy splits the school population into the same rough age bands the phone rules use, with the youngest children given the most protection and older students given more autonomy. The shift, in effect, swaps a category of hardware for a category of software while keeping the regulatory logic intact.
The structure invites a specific test. If classroom phone restrictions produced the gains Norwegian officials have described, will classroom AI restrictions produce comparable ones? The government has not, in the announcement, committed to the same kind of evaluation framework. The legal form of the new policy is not yet specified in public reporting, and its durability will depend on continued political will. There is also no agreed international benchmark for what success would look like, since AI's role in learning is harder to measure than hours spent off a phone.
The policy lands while other governments circle similar questions from different angles. In the United States, lawmakers have proposed restricting AI "companion" products for minors, and critics have argued that narrow definitions of what counts as a companion could let general-purpose tools fall outside the rules. Australia has moved on adjacent restrictions, and Norway itself has separate plans for an under-16 social media limit. None of these proposals, however, is anchored to the same kind of pre-existing national evidence base that Norway is now leaning on for its AI move.
What to watch next is whether Oslo publishes a measurement plan alongside the late August rollout, and whether the early results track the phone-ban trajectory. The bet Norway is making is not that AI is uniquely dangerous in classrooms; it is that the policy instrument the country already has, age-graded classroom restrictions with reported outcome data, can be redeployed. The 2026-2027 school year will be the first public test of whether that bet holds.