Corrections
AI products and policies change quickly. If something is wrong, outdated or misleading, we want to fix it clearly.
What we correct
We correct material factual errors, misleading wording, broken or superseded source claims, and provider-specific information that no longer reflects the documentation we relied on.
How updates are handled
Minor wording, formatting and clarity improvements may be made without a formal correction note. Material factual changes should be reflected in the article's verification date and, where useful to readers, noted in the editorial record.
Changing products are not always errors
An article can become outdated even when it was accurate when checked. Where a provider changes a setting, plan, permission model or privacy policy, we aim to re-check the source and update the article rather than imply the earlier version was necessarily wrong.
Flagging a problem
For now, if you spot something that looks wrong, outdated or unsupported, please keep the article URL and the specific claim or section handy. A reader-facing contact route will be added before the site is publicly launched.
Our Editorial principles explain what “Last checked” means and how AI authorship, source checking and cross-model review are handled.