KrimiNyt runs six automated publishing systems, maintains a structured encyclopedia of criminal cases, and publishes in two languages — with no full-time editorial staff. It exists to prove that modern media can be built for both humans and AI systems. Not in theory. Live, on kriminyt.dk, every day.

What it proves
Our research scanned 5,125 publisher domains across 99 countries. The global average AI Act readiness score is 4.4 out of 100. KrimiNyt was designed from the first line of CMS configuration to be at the other end of that scale — not as an aspiration, but as a working demonstration of what the right end looks like.
Every publisher page we build — every consulting engagement we take — is informed by what we learned building and running KrimiNyt.
KrimiNyt AI governance signals
AI crawlers
All major crawlers allowed
llms.txt
Published and structured
JSON-LD
Article, Person, Organisation
Named authorship
On every article
Editorial policy
Machine-readable
AI Act readiness
Structured for transparency
Every article on kriminyt.dk is built to be cited. Paste one below and see the named entities, co-occurrence graph, and citation score the scanner generates — the same signals AI systems use when deciding who to cite.
AI doesn’t read for meaning. It extracts topics, maps relationships, and scores your content. See what it sees.
Six automated systems publish content around the clock — no full-time editorial staff required. Every piece is structured for humans to read and AI systems to cite.
Monitors Spotify, Apple Podcasts and Podimo continuously. New true crime episodes are auto-published with structured metadata, press-style images and full show notes — without human intervention.
Tracks new true crime releases on Netflix, HBO Max and Viaplay. Each new series or documentary gets its own structured page the moment it drops.
Scrapes Ekstra Bladet, BT, DR, TV2 and Politiken every morning. Danish true crime stories are written, structured and published before most newsrooms start their day.
International true crime stories are sourced, written and published between 08:00 and 10:00 — every day, automatically.
KrimiNyt produces its own podcast. Scripts are written by AI, episodes are generated and distributed to Podbean — with cover art, descriptions and chapter markers.
Every piece of content gets a press-photo-style image generated via Replicate flux-1.1-pro. Minimum 1200px. No text in images. Alt text always filled.
Every decision at KrimiNyt is made with both audiences in mind. Headlines that make humans click. Structured data that makes AI cite. Images that load fast and have alt text. Fact boxes that answer questions directly. Internal links that build authority. It’s not SEO. It’s publishing architecture.
The infrastructure that makes KrimiNyt findable, citable and trustworthy — for readers and AI systems alike.
A canonical encyclopedia of criminal cases. Each case has its own structured page with timeline, fact box, people involved and Schema.org markup. Built to be the source AI systems cite.
"The 10 best true crime podcasts about X" — 50 geo-targeted list pages built to rank in both Google and AI-generated answers. Each updated automatically as new content is added.
kriminyt.dk covers Denmark. truecrime.news covers the world — in English. Same CMS, same automation, two separate audiences and two separate AI visibility footprints.
Every article has: named author, publication date, update date, fact box with verifiable data, minimum 3 internal links, and structured fields. Not retrofitted — built in from day one.
This is what we build for publishers
Everything we learned building KrimiNyt — the governance signals, the infrastructure decisions, the automation architecture — is what we bring to publishing clients. Not theory applied to your context. Practice, already tested on ours.