Every closing with an existing mortgage runs through one document that nobody questions: the payoff letter. It says who gets paid, how much, and where to wire the money. It arrives by fax or email, it looks like every other payoff letter you’ve ever seen, and it directs what is usually the single largest wire in the file — bigger than the buyer’s down payment that gets all the wire-fraud press. I wrote about the down-payment side back in April. This is the other side, and it’s worse.

It’s worse for a specific structural reason: the payoff letter is the one wire instruction in the file that originates outside your control and outside your customer relationship. You know your buyer. You know your seller’s agent. You don’t know the loss-mitigation department of a servicer three ownership changes removed from the original lender. When a document shows up claiming to be from them, your basis for trusting it is that it looks right.

The Fraud Is Boring, Which Is Why It Works

Nobody forges a payoff letter from scratch. The criminals intercept a legitimate one — through a compromised email account, usually on the borrower or servicer side — and alter just enough to redirect the money. A routing number. A beneficiary name. The callback phone number, so that when your closer diligently calls to verify, the friendly voice confirming the wire details is the fraudster’s. Same logo, same layout, same legal boilerplate, same per-diem table. The letter is 99% authentic because it started life authentic.

And the industry’s defense against this is one person eyeballing the document and making a callback. That defense fails for reasons that have nothing to do with how smart or careful your closers are. It fails because the review is inconsistent — the letter that arrives at 4:45 on the day before closing gets a different level of scrutiny than the one that arrives with a week to spare. It fails because humans are good at spotting things that look wrong, and this document is engineered to look right. And it fails because the callback, the one control everyone trusts, is only as good as the phone number — which is printed on the document you’re trying to verify.

What Systematic Verification Actually Looks Like

The fix isn’t more vigilance. Vigilance doesn’t scale, and it degrades exactly when volume peaks. The fix is making the verification systematic — the same battery of checks on every payoff, every time, regardless of how busy the office is or how legitimate the letter looks.

Some of those checks are internal to the document. Does the per-diem interest math actually compute? Is the payoff amount plausible against the principal? Is the good-through date stale? Does the wire beneficiary match the lender on the letterhead, or is a national servicer suddenly banking through a credit union two states away?

The more powerful checks are external, because the whole point of the fraud is that the document is internally consistent. Does the routing number actually belong to the bank named on the letter — not according to the letter, but according to the Federal Reserve’s own directory? Does this letter match what this lender’s legitimate documents look like — the layout, the letterhead, the real phone numbers — across hundreds of previously verified examples? A letter that claims to be from a lender but doesn’t match the fingerprint of that lender’s genuine paper is the tell no single closing office can catch, because no single office has seen enough of that lender’s paper to know.

This Is Why We Built Sentinel Plus

Sentinel earned its name guarding county records against deed fraud. Sentinel Plus points the same vigilance at the closing table. It reads every payoff letter with AI vision — including the fax-of-a-photocopy quality that defeats normal OCR — extracts the wire instructions, verifies the routing number against the Federal Reserve directory in real time, runs the full battery of fraud checks, and compares the document against a nationwide library of verified lender fingerprints. It returns a verdict in seconds — PASS, REVIEW, or FAIL — with an itemized list of exactly what looked wrong and where.

Two things it deliberately does not do. It doesn’t replace your callback procedures — it strengthens them, by telling you which callbacks deserve extra suspicion and giving you an independently sourced number to dial. And it never approves, releases, or blocks a wire. A human makes the final call on every payoff, every time. What changes is that the human is no longer the only control — and every verdict is permanently recorded, so when someone eventually asks “what did you do to verify this payoff?”, the answer is a report, not a shrug.

Payoff fraud works because it targets the one document everyone trusts and the one moment nobody has time to check. Make the check take seconds and run on every file, and it stops being a profitable crime.

See how Sentinel Plus verifies every payoff →