EDITORIAL POLICY

How Korea Signal makes editorial decisions

Evidence before volume

Zero posts on a given day is a normal outcome. We do not publish to satisfy a schedule. A candidate needs adequate evidence, source diversity, relevance beyond Korea, and original analytical value.

Fact and analysis

Articles distinguish observed facts and source evidence from Korea Signal's analysis. Unknown values remain unknown. We do not infer a Korea-to-global gap when comparable global data is unavailable, and we do not use false precision for qualitative evidence.

Sources and originality

Primary public data, corporate filings, official statistics, and official announcements are preferred. Reliable reporting may support factual context. Community and social signals may help discovery but do not stand alone as factual evidence. We do not republish article text or unlicensed news images.

Automation and AI

Software may assist collection, deterministic analysis, duplicate detection, planning, drafting, and quality checks. A language model may be used selectively when configured. Automated output is not treated as verified evidence, and no unreviewed draft can be published.

Human review

A human editor must approve every public article after reviewing its claims, sources, original contribution, similarity checks, policy risks, and rendered preview. Technical safeguards also block fixture data and drafts that fail a critical gate.

High-risk topics

Politics, elections, crime, accidents, war, death, medical claims, personal data, and misinformation-prone breaking news are manual-only or excluded by default. Korea Signal focuses on durable trends rather than breaking-news traffic.

Visuals

Actual metrics are presented with programmatic HTML, CSS, SVG, or charts. AI-generated editorial illustrations may be used selectively, labeled where needed, and never presented as documentary evidence of a person, product, place, facility, or event.