Google's share of what Norwegian 9 to 18-year-olds use for search fell from 72 to 47 percent in two years, and an NTNU review of 173 studies says nobody knows what that does to how they think

The 9 to 18-year-olds whose search behaviour becomes the market inside a decade are already spending most of their questions somewhere other than Google. Norwegian SciTech News, the English-language outlet of the Norwegian University of Science and Technology, published two figures together on 22 September that make the shift concrete.

The first is a survey from the Norwegian Media Authority. Google search use among 9 to 18-year-olds in Norway fell from 72 percent in 2024 to 47 percent in 2026, and 39 percent of 11 to 12-year-olds now use AI. The second is ung.no, Norway's public information site for 13 to 20-year-olds. Visits dropped from 24 million in 2024 to 18 million in 2025, a decline the outlet attributes to teenagers switching to AI tools for answers and to Google's own AI summaries.

The review sitting beside those numbers

Alongside the survey, Hilde Storrøsæter, a PhD research fellow at NTNU's Department of Teacher Education, published a review of 173 international research articles on the use of generative AI in schools, done with colleagues at Czech universities. The review covers work from 2022 to spring 2025.

The headline finding is a knowledge gap. In 80 percent of the studies, the participants were university or college students, so almost nothing has been measured on the age group whose Google share just collapsed. Most of the remaining studies rely on self-reporting rather than observing changes over time, and the geographic mix is thin: 28 studies from South, Southeast and East Asia, nine each from Spain and the UK, 15 from the United States, almost nothing from Africa, South America or Central Asia.

What the review says about outcomes

Where studies do exist, the pattern is mixed. Of 139 studies on the effect of AI on critical thinking, 67 reported only positive effects, 33 only negative, and 39 both. Of 103 studies on problem-solving, 57 were positive-only, 23 negative-only, and 23 mixed. Storrøsæter told Norwegian SciTech News that AI tools help when they are used to challenge and question, and hurt when they become a shortcut to a ready-made answer. She said her team was surprised the dual pattern came through so clearly.

What this changes for a marketing team writing for the next buyer

The Norwegian survey is one country, and the review says out loud that its own evidence base is thin. Even with that caveat, the direction is the direction other operator surveys have reported for the same age group in the same window: Google is one destination among several, and for many of the questions a page was written to answer, it is now the second one.

A team writing product pages, guides and answers today publishes for two audiences at once: a Google reader whose share is shrinking in the youngest cohort, and a ChatGPT, Gemini or Perplexity reader whose share is growing there. A page that only reads well as a search result loses the second audience, and a page that only reads well as an AI answer loses the first. The Norwegian numbers are a reminder that the split is not a projection for later this decade. In 9 to 18-year-olds in Norway it is already the majority pattern.

Source

Norwegian SciTech News: Young users ditch Google for AI, with unknown consequences, published 22 September 2026 by Sølvi Normannsen. Underlying survey by the Norwegian Media Authority. Literature review led by Hilde Storrøsæter, NTNU Department of Teacher Education. Hacker News discussion at item 49798451.