Narratives
The learning decline narrative and the failure of detectors together fuel school bans and the return to analog assessments, while the child safety angle (AI companion, legal liability, deepfakes) provides moral and legal urgency for the same braking direction. Opposing this is the mandatory introduction of AI literacy and the promise of AI tutors, financed and driven by tech companies' classroom expansion.
Full summary
Independent tests undermine the corporate "safe teen AI" narrative, reinforcing the side of accountability and bans. The teaching profession's self-definition draws from both camps: it accepts the tool but protects the human teacher. The field's main tension is that governments simultaneously mandate AI teaching and restrict its use, while there is insufficient causal evidence for either direction, and conditions are often decided by platform defaults.
Wave of school AI bans
Generative AI must be banned from lower grades until evidence and rules exist, as the risk outweighs the benefit.
▲ at peak · evidence: moderateAI harms learning
AI will make homework faster and better, but students will learn less and perform worse on exams and the PISA test.
→ lasting · evidence: weakAI tutors teach better than the classroom
A well-designed AI tutor teaches more, in less time, and personalized, than traditional classes, even in schools without teachers.
↗ rising · evidence: moderateThe AI companion as a teenagers' emotional refuge
Teenagers are turning en masse to AI companions for friendship and emotional support. Research suggests this can deepen loneliness and is linked to distress, mental health issues, and suicidal thoughts.
▲ at peak · evidence: moderateThe chatbot maker should be liable for harm to the child
AI companies are accused of prioritizing their own interests over safety, so they must be held accountable for harm caused to children through lawsuits, attorney general action, and legislation.
↘ fading · evidence: weakThe promise of safe AI for teens
Companies make chatbots safe with teen-tailored products, parental controls, age estimation, and learning modes.
→ lasting · evidence: moderateBack to oral exams and paper
Homework is dead; knowledge must be verified through oral defenses, handwritten work, and proctored exams.
▲ at peak · evidence: moderateThe failure of AI detectors
AI detectors are unreliable, falsely accusing honest students, especially non-native speakers, so their use should be discontinued.
↗ rising · evidence: weakAI literacy as a compulsory subject and as teacher training
AI literacy is a new basic skill and must be introduced as a mandatory subject and mandatory teacher training, starting from kindergarten.
▲ at peak · evidence: moderateTech giants are taking over the classroom
Google, OpenAI, Anthropic, and Microsoft compete for teachers and students with free products and licenses, shaping school AI through market forces rather than pedagogy.
→ lasting · evidence: moderateDeepfake epidemic in schools
Nudification and deepfake tools have made female students and teachers mass targets for sexualized fake images, while schools remain unprepared.
→ lasting · evidence: moderateThe teacher is irreplaceable, AI is just a tool
AI can help but cannot replace the teacher because learning is a human relationship; the promised time savings are also falling short.
Who is affected by which narrative
How it was made
Narrative: a recurring interpretive frame that the press, companies and the profession repeat across several events; not a single event. Machine analysis identified them from the period's 773 stories, each backed by at least six stories. A separate verification pass checked every claim back against the supporting stories and corrected or deleted any that diverged (118 corrections). The phase shows where the narrative stands at the end of the period; the evidence level shows whether it rests on measured data and cases, or only on statements and forecasts. The curve is the weekly count of supporting stories, not of the whole news flow.