How we choose and verify what we publish. In short: a named source for every fact (an official primary source, or - when there is no official announcement - an established outlet with direct access, clearly labeled), fact separated from read, corrections when we get something wrong.
Sources
We stand on official primary sources whenever they exist - blogs, press rooms, documentation, changelogs, regulatory publications and reports from the organizations themselves. We do not work with rumors or random aggregators.
When the source is a company, we treat it as a company source and watch for PR framing. On legal and regulatory topics we look for a regulatory primary source too, not just a company blog.
When a story has no official announcement, we use an established outlet with direct access to the facts and label it clearly as 'Media confirmation' in the sources block - so you can see what the source is, instead of it posing as official.
The three layers of every story
Fact - what was announced, by whom and when. Context - why it matters. Read - our interpretation, clearly separated from the fact.
Company numbers and claims are attributed to their source ('according to OpenAI', 'per Anthropic') and never presented as independently proven truth.
Copyright and trademarks
Facts and ideas by themselves are not subject to copyright; specific text, images and creative form are. That is why we work only with our own retelling, our own commentary and our own visuals, with a link to the source.
Trademarks, product names and logos belong to their respective owners; mentioning them is descriptive and implies no partnership or endorsement.
Corrections
When we get something wrong, we fix the text and describe the change on the Corrections page. Write to hello@wearecoded.com.
What we DON'T do
No rumors. No copying from the original (beyond short terms and names). No third-party photos, logos or product screens. No hidden tracking. No absolute legal claims.