A fact canon is the maintained set of verified facts an AI is allowed to state about a brand. Each fact is proposed, confirmed by a human, and kept current, so generation draws on truths a person signed off rather than whatever a model assumed.
Why it matters
A single verified fact is useful once. A brand needs the whole set kept in one place, current and authoritative, so every product reads the same truth.
Facts also change. A brand rebrands, a claim is updated, a correction supersedes an old answer. Without a maintained canon, a stale fact lingers and gets repeated.
The canon is what an AI is permitted to assert. If a statement is not backed by it, the honest move is to defer, not to improvise a plausible answer.
How Jinn treats it
Fama keeps the canon as a ledger. A fact enters as proposed, a person confirms the real answer, and confirming it supersedes the prior version, so there is always one active truth per claim.
Active facts overlay generation as authoritative addenda, and a human-confirmed correction overrides the brand's extracted record. The original record is never overwritten, the canon sits on top of it, and once a claim is corrected it is not flagged again.
When you confirm a correction, it lands in the Brand Truth Ledger, and every future Fama asset is required to respect it. See it
Where this shows up
One maintained set of signed-off truths.
The canon is what an AI is allowed to say about a brand, kept current by the people who own it. See how Fama holds the line.