Audit Lineage & Draft Evidence | Forseti AI

When a regulatory inspector asks "what did the AI say about this?", the customer needs a defensible answer. A summary screenshot is not enough. A chat transcript is not enough. The inspector wants the lineage: which record the assistant was looking at, which regulatory packs it had access to, which tool calls it made, which citations came back, which artifacts it produced, who approved them, and when.

Forseti's audit lineage view assembles all of that automatically. Every chat, every tool call, every workflow run, every draft, every approval transition flows through the same immutable audit trail as the rest of the platform. The chain-of-custody anchor that protects ordinary platform events protects every Forseti event too. The tamper-evident proof tap is emitted nightly; auditors can verify the chain end to end.

The draft-evidence layer is the other half of the picture. AI-generated evidence is created as draft-only. The approval transition (moving evidence from draft_ai_generated to approved) is enforced as a human action; the AI cannot transition its own evidence. The transition is recorded in the audit trail with the approving operator and the time of approval.

The point is not that Forseti is always right. The point is that every claim is cited, every artifact is reviewable, every interaction is in the chain of custody, and the customer can hand over a defensible answer.