Where do these scripts come from?

They are recovered during federal enforcement actions — the April 2026 New York indictment of a 23-person ring is the most recent example — and through FTC consumer-fraud research. The scripts presented here are sanitized but preserve the structure scammers actually train on.

Are scammers really using written scripts?

Yes. Federal prosecutors recovered training documents during the New York indictment showing typed scripts, jurisdiction-specific procedure language, fake badge number ranges, and rebuttals for common victim objections. Modern scam rings operate like call centers with quality-control coaching.

Why do all the scripts have a "don't tell anyone" line?

Because it works. Isolation from the family verification chain is the single most reliable manipulation in the entire scam. Train every senior in your family that the phrase "don't tell anyone" — for any reason, in any context, on any urgent call — is itself the strongest possible scam signal.

How do scammers know my grandchild's name and details?

From public social media (Facebook, Instagram, TikTok), tagged family photos, comments on posts, and data broker websites that aggregate family relationship data. A single public Facebook post tagging your grandchild is often enough.

What's the single most powerful defense against every script?

A pre-agreed family safe word that any caller must provide before urgent action is taken, combined with a trained habit of hanging up and calling back on a saved number. These two defenses together defeat every variant of the script regardless of how convincing the AI voice clone is.

Grandparent Scam Script Examples: 3 Real Scripts Recovered From Federal Cases

Federal prosecutors and the FTC have recovered scam-ring training scripts during multiple enforcement actions, including the April 2026 New York indictment of a 23-person ring. The scripts are remarkably consistent because what works keeps working: panic, a familiar voice, a story, an authority figure, and an irreversible payment. Below are three real script patterns — sanitized to remove operative details but preserving the structure scammers actually use. Each line is annotated with the manipulation lever it pulls. Read these once and the pattern becomes unmistakable when you hear it.

Script 1: The Car Accident

[Cloned grandchild voice, crying]: "Grandma? It's me — please, please, I need your help." → Emotional anchor. Forces grandparent into "helping" mode before any verification.

[Grandparent]: "Honey, what happened?"

[Caller]: "I was in an accident. There was another car… the woman, she's hurt. The police are saying it was my fault. They tested me and… I'd been drinking, Grandma. I'm so sorry." → Shame-inducing detail (drinking) prevents the grandparent from wanting to involve the parents.

[Caller]: "Please, Grandma, please don't tell Mom and Dad. I can't have them find out like this." → "Don't tell" — the most reliable scam signal in the entire script.

[Caller]: "There's a lawyer here, he says he can help. Please talk to him." → Handoff sets up the authority phase.

[Fake "lawyer"]: "Ma'am, this is Attorney James Roberts. Your grandson asked me to handle this. We need $8,500 wired to the courthouse account in the next two hours, or he'll be held until Monday." → False authority + artificial deadline + irreversible payment.

Script 2: The Foreign Hospital

[Cloned grandchild voice, weak]: "Grandpa? It's me. I'm in Mexico. There was a bus accident — I'm in the hospital." → Foreign country + medical = harder to verify, easier to pressure.

[Caller]: "My passport got stolen, my phone is almost dead. I'm using the hospital phone. They won't release me without paying first." → Establishes communication scarcity, justifies "I can't talk again later."

[Caller]: "Mom and Dad are on a cruise — I can't reach them. You're the only one who can help." → Isolates from the verification chain.

[Hand-off to "doctor"]: "Sir, this is Dr. Mendoza at Hospital General. Your grandson has internal bleeding. We need a $12,000 deposit before we can begin emergency surgery. Time is critical." → Medical authority + life-or-death stakes + immediate payment demand.

[Doctor]: "Please wire it through Western Union to this name… I will stay on the line with you." → Irreversible payment + stay-on-line kill switch.

Reality check: No hospital, foreign or domestic, demands wire-transfer prepayment from a relative as a condition of emergency surgery. Foreign hospital systems work on insurance verification, embassy notification, and post-treatment billing — not panicked wires from grandparents.

Script 3: The Police / Bail Emergency

[Cloned grandchild]: "Grandma? Don't be mad. I got pulled over. They found something in my car that wasn't mine but they're saying it's mine." → Plausible plot + plausible denial. The "wasn't mine" framing makes the grandparent want to defend the grandchild.

[Caller]: "I have one phone call. Please don't tell Dad — he'll kill me. The lawyer said you might be able to help." → Don't-tell + hand-off setup.

[Fake "officer"]: "Ma'am, this is Sergeant Davis with the Sheriff's Office. Your grandson is under arrest. Bail has been set at $9,500. We can release him to his attorney's custody if bond is posted within the hour." → Specific (false) procedure language + artificial deadline.

[Officer]: "Court accepts wire to the bond office account. I'll stay on the line and walk you through it." → False payment method + stay-on-line.

[If grandparent hesitates]: "Ma'am, every minute we wait, the charge gets enhanced. He could be looking at a felony if we don't move fast." → Escalating consequence pressure to override doubt.

What Every Real Script Has in Common

Across thousands of recovered calls, the structure is the same:

Voice anchor — cloned voice of the grandchild creates emotional trust before logic engages.
Shame seed — accident, arrest, or DUI creates a reason not to involve parents.
Don't-tell instruction — explicit isolation from the verification chain.
Authority handoff — lawyer, officer, doctor adds legitimacy to the financial request.
Specific (false) procedure — invented court fees, hospital deposits, or bond accounts.
Irreversible payment — wire transfer, gift cards, crypto, never traceable methods.
Artificial deadline — "in the next hour or he goes to felony / dies / is held until Monday."
Stay-on-line kill switch — the authority figure offers to wait while the wire is sent, blocking callback verification.

Recognize any three of these in real time and you have identified the call as a scam.

The Counter-Script

Memorize and rehearse this single line:

"I am going to hang up and call my grandchild's saved number to verify. If you are real, you will not object."

Then hang up. Don't negotiate. Don't apologize. The objection that follows is the strongest possible confirmation that the call is fraudulent. According to the FTC, victims who hung up to verify were 80% less likely to complete a fraudulent transaction.

Practice Before the Real Call Arrives

Reading scripts is awareness. Hearing one is preparation. TrustboxAI runs a controlled, educational simulation that uses AI voice cloning so your family hears the pattern once in a safe environment. Research published in the Journal of Experimental Psychology shows experiential exposure increases threat recognition by up to 65% compared to reading alone. See also our 5 red flags of an AI cloned voice call and the family safe word setup guide.