I’ve spent twenty-six years building systems that interrogate documents. Is the acknowledgment good. Does the legal match. That machinery is the spine of this business, and it works. Which is precisely why the fraud I hear about most from independent shops walks straight around it.

The scale is easy to underestimate. NDP Analytics, working from ALTA’s 2024 member survey, found that 28 percent of title companies hit at least one seller impersonation attempt during 2023 — nearly three in ten. The more alarming number: 19 percent saw an attempt in April 2024 alone. A full year at 28 percent, then a single month at 19. Those describe very different threat environments, and nothing I’ve seen since that survey suggests the curve bent back down. Nobody forges anything here. Somebody picks a property, claims to be its owner, sells it, and takes the money. The deed is a real deed, and the name on it is the real owner’s name. It just isn’t their name.

Why It’s Almost Always a Vacant Lot

Think about what a fraudster needs: an owner who won’t notice. No tenant to see a sign go up in the yard, no mail arriving at the property, no utility account, no lender watching the tax bill through an escrow account. An unimproved parcel gives all of it at once. Better still, vacant land is often owned free and clear — no mortgage means no payoff, and no payoff means no third party who has to be told the property is selling. Add an owner three states away, or elderly, or dead since 2019 with an estate nobody has touched, and the parcel is effectively unwatched. Targeting is trivial, because we made the data public on purpose: anyone can pull a list of out-of-state owners of unimproved parcels in your county before lunch. The real owner finds out when they go to sell in 2029, or when a tax notice bounces — long after the money moved and kept moving.

Your Exam Is Answering the Wrong Question

Vacant land seller impersonation isn’t a failure of your examination. It’s a fraud specifically designed to pass your examination, and it does. The notarization is real — a notary genuinely witnessed someone present a credential, and in a remote closing that person answered the knowledge-based questions off the true owner’s public history. The chain is spotless, because it’s the true owner’s chain. Every control in the workflow is built to validate a document, and this fraud hands you documents that are perfectly valid on their face. What catches it is a different question entirely. Not “is this instrument good,” but: is this human the person they claim to be, and does the shape of this transaction make sense for that person. Neither is a curative question, which is why curative doesn’t answer them.

Any One Is Innocent. Four Together Is a Shape.

The markers are not individually suspicious. Unimproved parcel. Owner of record’s mailing address far from the property. A seller who insists on handling everything remotely and won’t get on a phone. Priced to move. All cash, no financing contingency, which removes the lender and the appraiser along with it. Proceeds directed to a name or a bank that doesn’t match the seller. Urgency with no business reason behind it. Take any one in isolation and it’s Tuesday. Plenty of legitimate people inherit dirt in a county they’ve never visited and want it gone by Friday. But these markers aren’t independent variables, and you shouldn’t score them like they are. Three or four stacked on one file is not four coincidences. It’s a shape, and the shape is the signal.

The Answer Is a Rule, Not a Hunch

You cannot solve this by asking examiners to develop an instinct. Some will, and the good ones already have — worth nothing on the Thursday your best examiner is out and the file lands with someone in their second year. What you need is a written rule: when this combination of markers appears, the transaction does not close until someone reaches the owner of record independently, through a channel the transaction did not supply. That last clause is the whole thing. The email address came from the fraudster. So did the phone number, the “attorney,” often the agent’s only contact for the seller. Every channel inside the deal is one the impostor controls. Independent means the mailing address on the tax roll, the address on the prior deed, a number you went and found — slow, manual, and the one step the fraud has no answer for. And the rule has to fire on its own, off the file’s attributes.

Yes, This Creates Friction. Take It Anyway.

The objection is real: this puts work on transactions that are overwhelmingly legitimate, and it lands hardest on deals pushing to close fast. Somebody in your office will say we’ll lose business, and they won’t be entirely wrong — it’s a thin slice of your volume, which is the point of requiring several markers, but it isn’t zero. You do it anyway because the cost is wildly asymmetric. Delaying a clean $80,000 lot by three days annoys a real seller, who then closes and forgets about it. Getting one wrong means the policy amount, plus whatever the innocent purchaser built on the land, plus litigation with a furious true owner — a loss that is functionally unrecoverable, because the funds were converted before your file was even scanned. And be honest about the alternative: the seller pushed back on verification, so we let it go. That isn’t a shortcut. It’s the mechanism. It’s how every one of these closes.

Where the Software Earns Its Keep

This is a detection problem before it’s a judgment problem, and detection is what machines are good at. Sentinel watches a file’s attributes as they arrive — the property, the parties, how the money is moving, the tempo of the deal — and recognizes when enough have stacked up to be worth a human’s attention. It doesn’t decide whether a transaction is fraudulent. It makes sure the pattern reaches someone who can decide, while the file is still open and the wire hasn’t gone out. The failure here is almost never that somebody looked at the shape and got it wrong. It’s that nobody was prompted to look. Documents have gotten harder to fake and identities have gotten easier. Our defenses were built for the first problem. Fix the process now, while it’s still a rule you chose rather than one an underwriter hands you.

See how Sentinel flags absentee-owner and vacant-parcel risk →