Norway will ban children aged 6 to 13 from using AI tools in school from late August. Teenagers aged 14 to 16 can use them only when a teacher is present. Students 17 and older are expected to learn to use AI properly on their own, as preparation for work and further study.
The government's argument is simple. Using AI to write an essay or solve a problem means the child never actually learns how to do it. Prime Minister Jonas Gahr Støre put it plainly at a press conference: the most important thing school should do is teach children to read, write, and do mathematics. Letting AI do those things for them defeats the point.
This follows a pattern Norway has already tested. The country banned smartphones from schools in 2024, driven by falling national test scores. A study by the Norwegian Institute of Public Health, which tracked more than 400 middle schools, found the ban cut bullying by around 46% among girls and 43% among boys. Girls also visited mental health specialists far less often after the ban came in, and their grades improved. The results were strongest for girls from lower-income families. That evidence gave the government the confidence to keep going.
On the same day as the AI announcement, the government said it would also introduce legislation to put more physical books back into classrooms. Norway had been shifting toward tablets and screens since the iPad arrived in 2010. That direction is now being reversed.
Norway is also preparing a social media ban for all children under 16, with a bill expected to go to parliament before the end of 2026. Australia introduced a similar under-16 social media ban in late 2025. The UK is moving in the same direction.
The pattern is not limited to Europe. China's Ministry of Education issued guidelines in 2025 barring primary school students from using open-ended AI tools on their own. In the US, a bipartisan bill called the GUARD Act passed the Senate Judiciary Committee unanimously in April 2026. It targets AI tools that simulate emotional relationships or sustained companionship with minors, requires age verification, and sets criminal penalties for companies whose tools encourage self-harm among children. The bill still needs a full Senate vote.
There is a genuine enforcement question underneath all of this. Research from the US, published in early 2026, found that AI use among teenagers in schools with bans is virtually identical to use in schools without them. Over 40% of surveyed American teens said most of their friends use AI for schoolwork even when they are not supposed to. A ban in school does not stop a child from using ChatGPT at home on a phone.
That gap matters for anyone running a business that touches schools, children's services, or family-facing products. Regulatory pressure on AI and children is building fast across multiple governments at once. The specific rules vary, but the direction is consistent: age limits, parental controls, supervision requirements, and liability for companies that get it wrong. Businesses in education, edtech, media, retail for children, and consumer services should assume this regulatory environment will get tighter, not looser, over the next two to three years. Norway's move today is one data point in a much larger trend.