This text is the English version of the opinion piece in the online newspaper Nationen. The text is written by my colleagues Gilda Seddighi, Rebecca Lynn Radlick (Norce), Hilde G. Corneliussen (both Vestlandsforsking), and me.
Every day, students use AI for schoolwork, to grasp difficult subjects, and to complete homework. In public discourse, this use of AI is most often associated with cheating, taking shortcuts, or a lack of personal effort. This perspective is too narrow.
The debate surrounding AI in education has focused on rules and guidelines. In June, the government tasked the Norwegian Directorate for Education and Training with developing national recommendations for using AI in primary and secondary education. The recommendation is to use AI as little as possible among the youngest age groups, while older students should receive support in using AI appropriately. The concern is that the uncritical use of AI might lead students to skip essential stages in the learning process. The Standing Committee on Education and Research has reached a consensus on measures to combat AI-assisted cheating on tests and exams. Determining the appropriate age for students and establishing what is permitted—and what is not—are important questions. However, all these proposals presuppose the presence of parents and teachers who can assist and guide the students.
However, not all children and young people have someone they can turn to. In a research project, we examined how social inequality affects young people’s ability to navigate an increasingly digitalized everyday life. We interviewed 27 young people, aged 16 to 29, with an immigrant background who are neither employed nor in education. Most of them use AI regularly—for everything from learning about topics that interest them to sharing secrets. They use AI not because they are fascinated by the technology itself, but because they lack people in their lives they can ask for help or speak to in confidence.
For young people who have adults around them whom they can ask about matters both large and small, AI serves as an additional tool—helping them get started more quickly or articulate their thoughts more effectively. If they have a question, they have someone to turn to for help.
However, for some of the young people we met—those who lack both close personal relationships and contact with public support services—AI becomes something else entirely: a substitute for real, flesh-and-blood sources of support. One young woman told us that she messages the AI bot on Snapchat when she is feeling sad or worried and does not feel safe enough to reach out to anyone else. She doesn’t believe it actually helps, but, as she put it, “at least they answered me.” AI is always at one’s disposal, ready to compensate for a lack of emotional and practical support. It is simply a matter of using the right prompt.
One drawback, of course, is that AI cannot offer the same things as real human beings. Responses from a chatbot must always be evaluated critically and adapted before they can be useful. A young man we interviewed uses ChatGPT to help him understand letters from NAV and determine what rights he has. He has discussed his situation with friends and family, and can talk through with them how he intends to respond to NAV. Another young man comes from a family that has moved many times in search of work, and his parents have held only temporary jobs.
After five years without a job, he has lost both friends and his trust in NAV (the Norwegian Labor and Welfare Administration). To navigate the job market and make a living from his 3D design skills, he asks ChatGPT which niche to target and, “What kind of people should I reach out to?” He checks several websites to verify the answers. Ultimately, however, he decides to follow and engage with influencers instead, as he can’t determine whether ChatGPT’s information is reliable.
Young people with a support system can use AI as a supplement in collaboration with an adult. Those who use AI as a replacement, however, must do so alone or without adequate support. Consequently, they are also left with the arduous task of scrutinizing the chatbot’s output to ensure its accuracy.
This is what we mean when we say that AI strategies cannot be solely about AI. How young people use AI can sometimes reveal where they lack support—from adults, family, friends, or professionals. This might mean a teacher with time to explain things, a caseworker at NAV who can provide follow-up, an adult at a youth club, or a librarian. What is needed isn’t just new technology rules, but more adults and trusted people to turn to.
The same applies in schools; it is more about access to adults and teachers than about using AI as a shortcut in the learning process. When a student gets stuck on an assignment, and neither the teacher nor anyone at home has time to help, AI often becomes the only available resource. If students increasingly turn to AI because they aren’t getting the individual follow-up they need, the discussion must also address teacher-to-student ratios and the availability of adult support—not just guidelines for using technology.
We need rules for AI in schools, and we need to equip students for an AI-driven future. But we also need politicians who are genuinely troubled by the fact that some young people feel so alone and isolated that they turn to AI to make sense of the world—for how can we ensure the safety of those who use AI alone in their beds to seek answers? When we adopt new technology, it is in people that we must invest. Therefore, AI strategies cannot be solely about AI.
Leave a Reply