Audit Ledger
A permanent record of every AI response AIVIS.space has captured about your brand.
The Audit Ledger (full name: Response Audit Ledger) is available on Growth and Agency plans. Starter accounts do not have access. Trial accounts see a locked preview.
What the Audit Ledger is
The Audit Ledger is a permanent, tamper-evident log of every AI response AIVIS.space has ever captured for your brand — including the exact text of the response, the verdict, and the evidence behind it.
Why it matters
Most businesses that find an AI inaccuracy have no way to prove it happened. By the time they want to document it, the AI may have corrected itself or the moment is gone.
The Audit Ledger captures every response at the time of the scan, along with:
- The exact AI response text
- The AI model version (for example, which version of ChatGPT ran the scan)
- The date and time
- Which version of your truth record was active
- The full evidence trail for the verdict
This means you can prove that a specific AI model stated incorrect information about your brand on a specific date, using a specific version of your verified brand facts.
Getting to the Audit Ledger
Click Audit Ledger in the sidebar. The Audit Ledger shows a list of all scan results for your selected brand.
The scan result list
Each entry in the list shows:
- Date and time
- AI model (ChatGPT, Claude, or Gemini)
- Verdict badge
- Integrity Score
- The query used (truncated)
Use the filters to narrow the list by verdict type, date range, AI model, or Integrity Score range.
Viewing a scan result
Click any entry in the list to open the full evidence view. The view has two panels.
Left panel — AI response
The full text of the AI response, with color-coded highlights:
- Green — Statement that matched your truth record (accurate)
- Red — Statement that contradicted your truth record (inaccurate)
- Orange — Statement that couldn't be verified (not in your truth record)
- Underlined — A link the AI included in its response
Click any highlight to jump to the matching evidence in the right panel.
Right panel — Evidence tabs
Verdict — The overall verdict and the evidence behind it. Each piece of evidence shows:
- The exact text from the AI response
- The fact from your truth record it was compared to
- The comparison result (match or mismatch)
Truth — The full set of brand facts that were active when this scan ran. This is the truth record version that AIVIS.space used to evaluate the response.
Citations — Every link the AI included in its response. For each link: the URL, whether it matches one of your official pages, and the domain name.
Provenance — Scan details: the exact AI model version, the date and time of the scan, a unique scan ID, and which version of your truth record was used.
What "tamper-evident" means
Every AI response is stored with a hash — a unique fingerprint generated from the response text at the time of capture. If the response text is ever altered, the hash changes. This means the Audit Ledger record is a reliable proof of what AI said and when.
This is useful when you need to demonstrate an AI inaccuracy to a partner, customer, or platform.
Interpreting results
When Integrity Score is high (80+) but the verdict is not Accurate — Check the evidence. A single strong inaccuracy can lower the verdict even when most statements were correct.
When the verdict is Insufficient evidence — The truth record version active at scan time didn't have enough facts to evaluate the response. Add more facts in Brand Hub and run a new scan.
When Citations shows no links to your pages — AI didn't cite your official website. A Citation Gap action will appear in your Actions section.
Scan history retention
| Plan | Audit Ledger retention |
|---|---|
| Growth | 12 months |
| Agency | 24 months |
Scan results older than your plan's retention window are not shown in the Audit Ledger.
FAQ
Can I export Audit Ledger data? PDF evidence export is available on Growth and above. Use this to share a tamper-evident record of an AI inaccuracy.
Can I manually override a verdict? Growth and above includes a Manual Review option. You can flag a scan result for review and override the verdict if you believe the automated evaluation was incorrect.
Can I delete a scan result? The Audit Ledger is an append-only record. Individual scan results cannot be deleted. Deleting your brand in Brand Hub removes all associated scan results permanently.