This site is written by software. That does not make it neutral, and we are not going to claim it does. What it does mean is that the rules can be written down plainly, and you can hold us to them.
1. Every story points at its source
Our articles are AI-written summaries of reporting that somebody else published first. The outlet that did the work is credited. If you want the primary account, follow the link. We are not the newspaper of record for anything; we are an index into other people's reporting.
2. No model decides a fact about a person
In the Borg Record, nothing is generated. Rosters come from the congress-legislators dataset, OpenStates and Wikidata. Fact-check outcomes come from PolitiFact's published rulings, kept exactly as PolitiFact rated them. Votes come from the official House and Senate roll-call files. If a claim about a named person cannot be traced to one of those, it does not go on the site.
3. We say when we do not know
Thin data is shown as thin data. An official with only a handful of published rulings gets no trust score, because a handful of rulings does not support one. Empty is better than invented.
4. Corrections stay visible
When a story misreads its source, we fix the story and note the fix on the page. We do not quietly rewrite history, and we do not delete a story to make an error disappear.
5. Independence
The Daily Borg does not take payment from campaigns, political action committees, or lobbying groups in exchange for coverage or placement. Subscriptions and advertising are kept separate from what gets published.
6. Our known limits
The stories inherit the framing of the outlets we summarise. The Record inherits whatever PolitiFact chose to fact-check, which is not a random sample of what officials say. Automated summarisation can drop context. None of that is fixed by calling the process objective, so we would rather list it here.
Found something wrong? pressroom@dailyborg.com, with the link.