Back to Trust Center
Corrections log · public

Every correction we have ever published.

We never silently edit. Every correction — to an AI output, a source label, a factual error, or in response to a takedown — is logged here with a date. Subscribe to the RSS feed and audit us in real time.

  1. 01Other

    Corrections log goes live

    Today we published the Prism corrections log at /trust/corrections. From this point forward every correction we make to AI-generated content, source labels, attribution, or any other public artifact will be logged here with a date. We never silently edit.

  2. 02AI output

    AI disclosure badges added across all LLM surfaces

    We added an inline 'AI-assisted' label to every panel where GPT-4o-mini outputs are shown (Common Ground, Flip the Script, Perspective Summary, Front-page Common Ground band). This is not a correction of an error — it is a transparency upgrade. If you find an AI output that looks wrong, email hello@prismnews.io and we will log the correction here.

Spotted an error?

Email us with the story or source, the error you found, and a link or screenshot. We log every report within 7 days, fix verified errors in place, and add a dated entry here.

hello@prismnews.io