Editorial standards and AI methodology
We publish reporting assembled from named, checkable sources. This page explains exactly how that work happens.
How stories reach us
We monitor permitted feeds from wire services, government portals and institutional newsrooms. Nothing that arrives through ingestion is published automatically: incoming material is grouped into story clusters and scored for relevance before a journalist looks at it.
Cross-checking before drafting
Each cluster is broken into individual claims. A claim is only marked supported when independent sources agree. Where sources disagree, we record the disagreement and report it as disputed rather than choosing a side.
Where AI is used, and where it is not
AI assists with research summaries, first drafts, headline options and SEO packaging, always restricted to the source material we supply. AI never publishes, never approves its own work, and is never treated as a source. Stories drafted with AI assistance say so.
Human editorial review
Every story passes through an editor. Politically, legally, medically or financially sensitive stories require explicit editor approval, and our guardrails block publication when sourcing, attribution or image rights are incomplete.
Corrections
When we get something wrong we correct the story, keep the original wording on record, and publish the correction on the article and in our public corrections log.
Reader privacy
We collect only aggregate reading analytics needed to run the newsroom. We do not sell reader data, and newsletter consent is recorded with a timestamp and can be withdrawn in one click.