// AI threat desk · 15 Nov 2026
Home/Editorial policy and corrections

Editorial policy and corrections_

Readers act on what we publish, so every fact has to hold up.

Sources

Primary documents first: vendor advisories, NVD, CISA KEV, HKCERT alerts, court filings, regulator notices, company statements. A claim with a single source is attributed to it. A claim we cannot trace to anyone accountable is left out.

Naming victims

We name a company as breached only when the company itself, a regulator, a court filing, or two outlets with their own sourcing confirm it. Otherwise we write "a Hong Kong trading firm".

Use of AI

We use AI tools for research, translation and building fact tables. An editor checks the facts and sources of every story before it is published. AI-generated images appear only as clearly labelled illustrations.

Corrections

Errors are fixed as soon as we find them, with a dated correction note at the end of the story. We never silently rewrite. To report an error, email corrections@owleval.com.

Sponsored content

Sponsored and partner content is clearly labelled and its links are marked sponsored. Sponsors do not see or influence news coverage.