Editorial standards

What we publish, how we verify it and how we separate facts from alerts, analysis and opinion.

Facts, alerts, analysis and opinion

Every piece carries a visible label with its type:

• News: an event that has happened, reported without value judgements. It states what happened, where, when, who confirms it and what is still unknown.
• Alert: a warning, forecast or instruction issued by an authority. It describes what has been announced, not something that has already happened, and links to the original notice.
• Analysis: context based on data and linked sources. It separates data from interpretation and projection.
• Opinion: the author’s view, always signed. It is neither a verified fact nor the newsroom’s position.
• Only news stories can be published without a person approving each one: they are selected and written by an AI assistant under these same standards, and their public history says so. Alerts, analysis and opinion are always approved by a person.

Verification

Each story shows one of these statuses, explained in how we verify:

• Confirmed: publicly confirmed by the responsible organisation, company or authority.
• Corroborated: supported by at least two mutually independent sources.
• Developing: first details from an identifiable source; the story will be updated.

Sources

• We prefer primary sources: operators, public bodies, emergency services, incident response teams (CERTs), regulators, scientific papers and direct statements.
• Several outlets repeating the same wire story or press release count as one source.
• We do not attribute a cyberattack to a specific actor unless the attribution is confirmed: we say who makes the claim and on what evidence.
• An expert’s credentials support their expertise, but do not turn their predictions into facts.

No unproven connections or trends

• We do not suggest two incidents are related unless a reliable source demonstrates it.
• Coincidence in time does not prove a cause.
• A single incident is not a trend. We only speak of trends backed by data series or comparable studies, and we say so.

Headlines and language

• The headline states the main fact accurately. We do not write headlines as questions or hide what happened to bait clicks.
• We avoid alarmist language and exclamation marks. The system flags expressions that exaggerate danger, and a story containing them is never published automatically.
• We report real risks precisely, without exaggerating danger to attract readers.

Images

We only use images relevant to the event, with visible credit and licence. We do not use images of other events or illustrations that could be mistaken for photographs of the event.

Languages

We publish in Spanish and English. Each piece has an original language and is machine-translated into the other; translations are labelled and link to the original. If a machine translation changes any figure, it is not published until the newsroom reviews it.

Updates and corrections

Updates change the “last updated” date. Corrections are flagged in the story itself and on the corrections page.