AI & Content Policy

Our pipeline discovers published announcements through approved RSS sources, classifies them, and prepares an original editorial summary of the source material. It then generates SEO metadata, selects or generates an illustrative image, and assigns categories, tags and internal links.

Articles produced this way carry the Blue Ocean Editorial Desk byline and a visible statement identifying them as source-based editorial summaries rather than independently reported stories.

Automated articles must not invent facts, quotations, statistics, people, companies or sources. They must not add claims that are absent from the source material.

Where a claim cannot be matched to the source, the claim is removed and the item is routed to human review. Where risk indicators are present, the item is rejected.

We never state that automated content has been independently verified when it has not. Verification means consistency with the named source, not independent confirmation of the underlying facts.

Editors can require human approval before any automated article is published, and that setting is on by default. Editors can edit, hold, reject or unpublish any item at any time.

Every automated article retains an internal provenance record: source URL, source headline, source publisher, source and retrieval timestamps, generation time, model identifier, verification status and editor approval.

We use a source image only where use is permitted and attribution requirements can be met. Otherwise we use a licensed asset or an AI-generated illustrative image, clearly captioned as illustrative.

We do not generate imagery designed to imply that a specific real-world event was photographed.