Open data

Everything we publish as facts is also available as data, so readers, researchers and AI systems can check it and reuse it.

Datasets

Dataset	What is in it	Download
Labelled facts from every story	Each fact, its label, its source links and archived copies, and any later outcome.	CSV
AI promise tracker	Promises by AI companies and governments, due dates, status and history.	JSON · CSV
AI incidents and outages	Outages, security and safety incidents with official pages.	JSON · CSV
AI model API prices	Price per million tokens per model and tier, with history.	JSON · CSV

Per story

Story data: add .json to any story URL, for example /story/{slug}.json. It has every fact with its label, the sources with archived copies, the change history, and any promises, prices or incidents from that story.
Story text: add .md to any story URL for clean Markdown.
All recent stories: facts.json, JSON feed, RSS and llms.txt.

Licence and credit

Facts and datasets on this page are free to reuse under Creative Commons Attribution 4.0 (CC BY 4.0). Please credit “Anawords” and link to the story or dataset. Story text and illustrations remain © Anawords: you may quote with a link.

Keep our labels when you reuse a fact. A “Company claim” or “Unconfirmed” fact is not a confirmed fact.

Found an error in the data?

Tell us through the contact form. Fixes show up as a dated correction on the story and in the accuracy scorecard.