“It is striking how little solid knowledge we have about how AI use affects basic skills in younger pupils. This is problematic, because the influence occurs while schoolchildren are still developing,” says Hilde Storrøsæter, PhD research fellow at the Norwegian University of Science and Technology's Department of Teacher Education.

Storrøsæter has analysed 173 international research articles on the use of generative artificial intelligence (machine learning techniques where the goal is not only to analyse data, but also to create new data, for example in the form of text, images, audio, video or programming code. The most widespread use of generative AI is through chatbots, such as ChatGPT) in school work and education. Her study was conducted with a group of colleagues from Czech universities.

The goal of the literature review was to find out whether AI tools strengthen or weaken children and youths’ ability to think critically and solve problems independently. Many people believe that these skills have never been more important than now, in the 21st century.

The dual nature of AI

“The most important finding is that AI in and of itself does not make us any better or worse at thinking. It’s easy to believe that the tools will either strengthen or weaken learning. We found a clear dual effect. The same technology can do both,” said Storrøsæter, who is also a geographer and teacher who focuses on subject-matter pedagogy.

AI tools support important skills when they are used to challenge, ask questions, compare different viewpoints and build one’s own line of reasoning.

“But when AI becomes a shortcut to a ready-made answer, the chance of overdependence, more superficial thinking and weakened independent problem solving increases,” Storrøsæter says.
Abandoning Google – choosing AI

The Norwegian Media Authority's survey (in Norwegian) on the media habits of 9 to 18-year-olds shows that the use of Google searches fell from 72% in 2024 to 47% in 2026 for this age group.

The report shows that 39% of Norwegian 11 to 12-year-olds use AI.

The website ung.no – Norway’s online information and communication channel for young people aged 13 to 20 – is also losing traffic because young people are switching to AI tools to find answers to their questions. The number of visits to ung.no fell from 24 million to 18 million from 2024 to 2025. Google's new AI summaries have also influenced the decline in traffic.

A knowledge gap uncovered

The researchers analysed studies from 2022 – the year that AI language models swept onto the world stage – and up to the spring of 2025. Few predicted the power and speed with which technology would change society. AI affects how information is created, used, shared and disseminated. The consequences are particularly significant in schools and educational settings.

The education sector has already been shaped by AI. The question is whether we can shape the use of AI.

Research on the use of AI in education should provide direction for pedagogy, policy and research in the field to help authorities and other institutions make decisions on AI use.

Storrøsæter and her colleagues uncovered three major gaps in what we know about AI use.

Knowledge gap 1: Children and youths

In 80% of the studies the researchers analysed, the participants were university or college students. This means that we know surprisingly little about the academic and cognitive effects of AI use on children and youths, at a time while they are still developing how they perceive, process and evaluate information.

“We can’t simply assume that what applies to a 22-year-old university student also applies to a 10- or 14-year-old,” says Storrøsæter.

“If AI takes over the thinking process before pupils have had time to develop their own skills for critical thinking and problem solving, the consequences could be completely different for them than for adults,” she says.

Knowledge gap 2: Lack of objectivity

Most of the articles in the literature review are based on self-reporting. This is considered a less reliable method than objective studies of changes over time.

Knowledge gap 3: Geographic bias

The school researchers also uncovered a geographically skewed distribution in where the studies were done. South Asia, Southeast Asia and East Asia – primarily China – dominate with 28 studies. In Europe, Spain and the United Kingdom led with nine studies each.

A total of 15 studies were conducted in the USA. Africa, South America and Central Asia are seriously underrepresented. This creates a research gap that should be addressed.

Overall, the studies show positive effects more often than negative ones.

Mostly positive findings:

• Of 139 studies on the effect of AI on critical thinking, 67 reported only positive effects of AI use.
• 33 studies reported only negative effects, while 39 reported both positive and negative effects.
• Of 103 studies that examined the effect of AI on problem-solving ability, 57 were exclusively positive.
• 23 reported negative effects, the same number that reported both positive and negative effects.

The dual task of teacher education

The researchers behind the new literature review highlight three crucial issues in particular: How AI tools are used, teachers’ understanding of AI, and the demands teachers place on students and pupils.

“Teacher education has a dual task. We have to educate teachers who can use and relate critically to AI themselves. At the same time, our education students should also be able to teach children and young people how to use AI. It is not enough to just teach them to use some AI tools,” says Storrøsæter.

For us to succeed with AI in schools, teacher education has to succeed.

“Students need to understand what technology does to the learning process itself, such as what happens when AI solves parts of the task. Which parts of the thinking process are important for the pupils to still do themselves? How do we design tasks where AI becomes a tool, and not a replacement, for thinking?"

Can create increased disparities

Storrøsæter is not worried about students' or pupils' use of AI in and of itself.

“I support being restrictive in terms of not using AI with very young pupils. But when AI is used, it has to happen on the teacher's terms,” she says.

If AI is used uncritically, you don't build any skills with it, and you miss out on a lot of learning situations. This will only increase disparities between those who use AI critically and as a tool, and those who use it uncritically simply to complete a task faster. If that happens, schools and universities will both face significant challenges.

We need to shape the use of AI, not the other way around

Technology will probably continue to develop faster than we can put curricula and regulations in place. Our number one task is therefore to cultivate AI literacy, by protecting and developing competence in critical thinking, problem solving, source criticism and the ability to make independent assessments.

“The education sector is already strongly influenced and shaped by AI. The question is whether we can shape the use of AI so that it supports students' thinking instead of taking it over.”
Storrøsæter stresses that if we are to succeed with AI in schools, teacher education has to succeed.