Where stories come from
We start from published news reports from established outlets. A language model drafts each story from those reports, and a second check compares every claim with the source text. Claims we cannot support are left out. The outlets are linked at the bottom of each story.
Where the data comes from
- Wikipedia (CC BY-SA 4.0): descriptions and background text.
- Wikidata (CC0): structured facts and identifiers.
Wikipedia text is used under CC BY-SA 4.0 and Wikidata under CC0. Entity pages can lag behind real events, since they refresh on a schedule.
Limits
AI can still get things wrong. No person reviews every story by hand. If we get something wrong we correct it in place and list it on the corrections page. More about us is on the about page.