The standards we hold ourselves to, in public.

Two independent sources for any contested factual claim. Single-source stories require Managing Editor sign-off and are labelled as such internally.

Anyone criticised gets a genuine opportunity to respond, with a stated deadline, before publication. Non-response is stated in the piece.

Granted only when the information is important and obtainable no other way, and only with Desk Editor plus Managing Editor approval. The reason for anonymity is explained to the reader.

Published, dated, visible, never silently edited. A correction log is public. This is a trust-building instrument, not an embarrassment.

We do not pay for interviews. We do not accept payment, gifts or travel for coverage. Press trips and event access are disclosed in the piece.

Generative AI may be used for research assistance, transcription, translation drafts and idea generation. It may not be used to write publishable copy, to generate images presented as documentary, or to synthesise quotes. All AI-assisted work is disclosed internally in the Story Ledger. Any AI-generated visual is labelled to the reader.

All Originals, all Intelligence products and all cover-level features are fact-checked by someone other than the author before release. Live programming uses a pre-brief verification pass on all claims scripted in advance.

Written consent for contributors in documentary work, especially where subjects are private individuals. Music, archive and stills cleared before edit lock, not after.

No story is worth a staffer's safety. Risk assessment required for any reporting in conflict, security-sensitive or hostile environments, signed by the Managing Editor.