Use cases
The file came from the customer. Check it before you rely on it.
- Paystubs
- Appraisals
- Photos
- Estimates
01Where it can be faked
The customer sent the file. Here is what can change.
First · Mortgage
Income and asset docs, before you fund
You already promised the file met the guide. Income is the defect they cite most when the loan comes back.
The buyback starts the day you funded
Year-to-date and gross pay edited so the loan qualifies.
Wages and withholding altered so the numbers no longer add up.
Balances and deposits inflated to fake a down payment.
Then · Insurance
Jewelry to property claims
Jewelry appraisals and receipts. Property photos and estimates. They issue the policy, or pay the claim, on that file.
Treated as true at bind, or paid at claims
Value, name, or weight edited so the piece qualifies.
Amounts or dates altered on the receipt that supports the policy.
Damage added or erased to match the loss being claimed.
Line items and totals altered to inflate the payout.
02Inside the automation
Hidden instructions ride inside automation.
- Read-off text
- AI summaries
Hidden instructions planted in the file ride along in that text and steer the AI that reads it next.
A clean-looking summary carries a tampered source forward as if it were trusted.
03Expansion
The same check works on other packets.
Logistics
Bills of lading and proof-of-delivery photos before goods release.
Construction
Change orders and site photos before progress draws fund.
SIU review
Flagged files and the evidence packet before an investigator spends the hour.
On your files
See what your intake is already missing.
Bring a sample from live intake. We'll walk the edits and fakes your reviewers and automation miss, before they reach a decision.