How we work
Every story on this site was drafted with AI and approved by a human. Here is exactly what that means, what we won’t do, and how to tell us when we’re wrong.
The four labels
Sourcing
A news story needs at least two independent sources, or one primary source: the lab’s own announcement, the paper, the release notes, the filing. Syndicated copies of one press release count as one source, however many outlets ran it.
Quotes are verbatim from a source we link, or they don’t appear. We never compose a quote, never paraphrase into quotation marks, and never attribute a quote we couldn’t retrieve ourselves. A post we quote is shown on the page as we captured it, so it survives deletion.
Where AI is used
Our newsroom pipeline reads a few hundred source items an hour, groups them into topics, drafts the ones that clear the sourcing rule, and attaches the evidence for each claim. That drafting is done by AI, in our house voice, under rules that reject fabricated quotes, undefined jargon and hype.
Nothing publishes on its own. The drafting system cannot publish; that is enforced by the database, not by a policy. A human editor checks each claim against its excerpt, edits, and approves one exact revision. Rejections and edits are logged.
Illustrations are synthetic and labelled. We do not generate images that depict real people or real events; we render objects and metaphors, and say so in the caption.
Corrections
When we get a fact wrong, the story carries a red correction band at the top stating the wrong figure, the right one, the source and the time. Updates that add information carry a blue band. We don’t silently rewrite.
Spotted something? corrections@thesamplespace.com. We reply, and we fix it within the day if you’re right.
Independence
No advertising, affiliate links or sponsorships run on this site today. If that changes, they will be labelled at the point of appearance, never placed inside a Reported or Tested story, and this page will say so first.