Grandparent Scam

A grandparent scam is a fraud in which the caller claims to be a grandchild, or someone acting for one, and says that an emergency requires money immediately: bail after an arrest, a hospital bill after an accident, a lawyer's fee, a fine after a car crash abroad. The names differ by agency. The FBI's Internet Crime Complaint Center files it as "the Grandparent's Scheme" inside its Confidence/Romance Fraud crime type, which it defines as covering "any scheme in which the perpetrator preys on the targeted individual's 'heartstrings.'" The Federal Trade Commission's category is "Family Emergency Scams", which is the better description, because the relative in the story is not always a grandchild and the target is not always a grandparent.

What makes it its own scheme rather than a variation on impersonation generally is the combination of three demands made at once: act now, act alone, and act on a payment method that cannot be reversed. Each is doing separate work. Urgency removes deliberation. The instruction to keep it secret removes the single check that would end the call in thirty seconds, which is speaking to the relative who is supposedly in trouble. And the payment method, most often a wire, a courier collecting cash, gift card codes or a payment app, is chosen so that a later realization changes nothing.

The script anticipates the obvious objection, which is that the voice does not sound right. The traditional opening handles it before it is raised: the caller is crying, or says they were injured, or asks "do you know who this is?" and works with whatever name the target supplies. More recently the objection is handled by technology instead. The FBI's 2025 Internet Crime Report describes grandparent scams, which it also calls distress scams, as cases "in which voice cloning technology is used to mimic the sound of a loved one in distress", and reports that victims claimed losses over $5 million in 2025 to distress scams, with the pattern "evolving to mimic other family members or close friends in different types of emergency scenarios." How that technology works is a subject of its own; what matters here is that a familiar voice has stopped being evidence.

The second voice is a structural element, not a flourish. The FTC's own description names it: scammers "might pretend to be an 'authority figure,' like a fake lawyer, police officer, or doctor working with your family member." The hand-off does three things at once. It removes the weakest link in the performance, since the impersonated grandchild only has to speak briefly and under distress. It supplies procedural detail, a case number, a bond amount, a court time, which is the texture that makes a story feel real. And it converts a family conversation into a transaction with an official, which is a context in which being asked to pay quickly feels normal.

Secrecy is the request to interrogate, because it is the only one with no innocent explanation. Urgency has innocent explanations, and so does an unusual payment method occasionally. An instruction not to tell the rest of the family does not: the actual relative in genuine trouble has every reason to want their own parents told. The FTC's advice deals with it directly, saying to call someone else in the family "even if the caller said to keep it a secret", and especially when the person supposedly in trouble cannot be reached.

The verification question is the practical centerpiece and it is worth rehearsing before it is needed. The FTC suggests asking "a question only the real person would know the answer to, like 'What kind of dog do you have?' or 'Where did you spend Thanksgiving last year?'" Two features make it work. It is answerable instantly by the real person and not at all by a stranger reading a social media profile, and it can be asked without accusing anyone, which matters when the caller might genuinely be your grandchild. The stronger version is to hang up and call back on the number you already have, since a caller who controls the line controls everything you can hear on it.

Where the money goes is not incidental to the design. A payment made by the target themselves, on their own instruction, sits on the authorized side of the line that federal electronic-transfer rules draw, and the strongest reversal rights attach to transfers the consumer did not initiate. That single sentence is why speed of reporting matters more here than argument about who was at fault; the boundary itself is set out on our pages on fraud and on phishing.

The call runs a sequence, and every step is defensive.

1. Contact and identification. The caller opens in distress, or invites the target to supply a name. Details the family assumes are private, a school, a car, a recent trip, are frequently available online or from another relative's compromised account.
2. The emergency. An arrest, a crash, a hospital, a border problem. The common feature is that it explains why the relative cannot talk normally and why a third party is handling it.
3. The hand-off. A lawyer, an officer or a doctor takes the line and supplies procedure: a case number, a bond amount, a deadline.
4. The secrecy instruction. Framed as protecting the grandchild from embarrassment, from their parents, or from a judge who will be harsher if the matter becomes public.
5. The payment. A wire, a courier collecting cash from the house, gift card codes read out over the phone, or a payment app transfer. All are chosen for irreversibility.
6. The second ask, which is the tell most often missed. A first payment establishes that the target will pay, so a complication arrives: additional court fees, a fine, a medical charge.

A hypothetical example, and the arithmetic is the point rather than the amounts. A caller claiming to be a grandson says he was in a collision, is being held, and needs $8,500 in bail. A second voice, presenting as his lawyer, gives a case number and asks that the money be wired within the hour, adding that the matter is sealed and should not be discussed with his parents. The money is wired. Two hours later the lawyer calls back: the other driver is claiming injury, and $6,200 more is needed to keep the matter out of court. Total paid: $14,700 ($8,500 + $6,200).

Nothing about the second call is new information. It is the same story extended, and it happened because the first payment answered the only question the operator had: whether this person pays. A single ask that is never repeated is unusual in this scheme, which is why "I have already paid, so it must be real" is exactly backwards.

A grandparent scam has no upside, so what follows is what defeats it and what the usual defenses do not reach.

What genuinely reduces exposure

• Hanging up and calling back on the number you already have. A caller who controls the connection controls the voice, the caller ID and the second speaker on it.
• Asking a question only the real person could answer, which the FTC frames as the polite version: "What kind of dog do you have?" costs nothing if the call is genuine.
• Calling somebody else in the family despite the instruction not to. The secrecy request is the part of the script with no innocent explanation.
• Agreeing in advance, while nothing is happening, that no money moves in the household on the day it is first discussed. That defeats manufactured urgency without anyone having to identify the scheme in the moment.
• Reporting immediately if money has gone. Some wires and app transfers can be stopped in the first hours and almost none after that.

What the protections do not reach

• A familiar voice, which voice-cloning tools have made unreliable as evidence.
• Caller ID, which can be made to display any number or name.
• Details the caller knows about your family, which are frequently public or were taken from a relative's compromised account.
• A payment you sent yourself. Federal electronic-transfer protections are built around transfers you did not initiate, and this scheme is designed so that you do.
• Cash handed to a courier, gift card codes read aloud, and cryptocurrency, none of which has any reversal mechanism at all.