ShieldNook

Deepfake voice scams: the family code-word trick

Your phone rings. It’s your son, and he’s panicking: there’s been an accident, he needs money right now, and please, don’t tell anyone. The voice is unmistakably his. Except it isn’t.

Scammers can now use AI to copy a person’s voice from a short recording, like a video posted online, and use it to fake an emergency. The good news: one free, five-minute habit shuts most of these calls down. It’s called a family code word.

What is a deepfake voice scam?

A deepfake voice scam uses AI “voice cloning” to make a call or voice message sound like someone you trust: your child, a grandparent, your partner or even your boss. The scammer only needs a short clip of that person talking, which is often easy to find in public social media videos.

The FBI has warned that criminals use AI-generated audio to impersonate close relatives in fake emergencies and to try to get into bank accounts. The U.S. Federal Trade Commission (FTC) has issued the same warning about AI-boosted “family emergency” scams.

These calls almost always follow the same script:

  • A sudden crisis: an arrest, a car accident, a kidnapping or a medical emergency.
  • Extreme urgency: you have to act in the next few minutes.
  • Secrecy: “Please don’t tell Mom,” or “The lawyer says you can’t talk about this.”
  • Money that’s hard to get back: a wire transfer, gift cards, cryptocurrency, a payment app, or cash collected by a “courier”.

All four are designed to stop you from thinking clearly or checking with anyone else. That’s exactly what a code word fixes.

The family code-word trick, step by step

A code word is a secret word or phrase that only your family knows. If someone calls with an “emergency” and asks for money or personal information, you ask for the word. Your real family member knows it. An AI clone doesn’t. The FBI recommends exactly this: create a secret word or phrase with your family to verify their identity.

1. Pick a word that’s easy to remember but impossible to guess

  • Good: two random words that don’t belong together, like purple pancake or lighthouse banjo. (Pick your own, not these!)
  • Avoid: pet names, street names, birthdays, schools, or anything someone could find on your social media.
  • Make it something a panicked teenager or a grandparent will still remember under stress.

2. Share it in person

Tell everyone face to face, or on a call you started yourself. Don’t post it, don’t put it in a group chat, and don’t save it anywhere obvious. If someone needs to write it down, a password manager is the safest place.

3. Decide who’s in the circle

Include anyone who might ever call asking for help: kids, parents, grandparents, your partner, and anyone who looks after your children. Older relatives are frequent targets, so make sure they’re included and comfortable using it.

4. Agree on one simple rule

No word, no money. If a call or message asks for money, passwords, codes or personal details and the caller can’t give the word, you hang up. No exceptions, no matter how real it sounds.

5. Never give hints

If the caller says, “I can’t remember, it’s something about food, right?”, that’s your answer. A scammer will happily play along with any hint you give. Wait for them to say the word on their own.

6. Practice once, and change it if it’s ever used

Run a quick 30-second drill at dinner so everyone knows how it works. If the word is ever used in a real situation or shared outside the circle, choose a new one.

If you get the call: a 60-second plan

  1. Pause. The urgency is the scam. A real emergency can survive a two-minute check.
  2. Ask for the code word. Don’t hint. If they can’t say it, hang up.
  3. Call back on a number you already have. Use the contact saved in your phone, not the number that called you. Caller ID can be faked.
  4. Check with someone else. Call or text another family member or friend to confirm where the person really is.
  5. Don’t pay with gift cards, crypto, wire transfers or cash couriers for anyone you haven’t verified. Real police, hospitals and lawyers don’t ask for money that way.
  6. Report it, even if you didn’t lose money (see below).

Make your voice harder to clone

You can’t disappear from the internet, but you can make a scammer’s job harder:

  • Set social media accounts to private, especially ones with videos of you or your kids talking.
  • Think twice before posting long, clear clips of family members speaking.
  • Use your phone’s default voicemail greeting instead of recording one in your own voice.
  • Let unknown numbers go to voicemail. On iPhone, turn on Silence Unknown Callers in the Phone settings. Many Android phones have built-in spam protection and call screening in the Phone app, and most mobile carriers offer free spam-call filtering.

For small businesses: the same trick works at work

Voice cloning is also used to impersonate bosses and suppliers: “Please pay this invoice today, I’m about to board a flight.” Protect your team with a few simple rules:

  • Call back to confirm any request to send money, change bank details or share passwords, using a number you already have on file.
  • Require two-person approval for payments above a set amount, so no single phone call can move money.
  • Agree on a private verification phrase for urgent requests from leadership.

Where to report a voice-cloning scam

  • United States: ReportFraud.ftc.gov and the FBI’s IC3.gov.
  • United Kingdom: Report Fraud, which replaced Action Fraud, covers England, Wales and Northern Ireland. In Scotland, call Police Scotland on 101.
  • Canada: the Canadian Anti-Fraud Centre.
  • Australia: Scamwatch.
  • Anywhere: if you’ve already sent money, contact your bank or payment provider immediately. Acting fast can sometimes stop or reverse a transfer.

Your five-minute action plan

  • Pick a code word with your family tonight.
  • Share it in person, including grandparents and babysitters.
  • Agree on the rule: no word, no money.
  • Make social accounts with videos of your voice private.
  • Turn on spam-call filtering on your phone.

AI will keep making scams more convincing. The code word works because it doesn’t try to out-tech the scammers. It simply asks for something an AI can’t know. That’s the idea behind ShieldNook Superintelligence .si: simple, smart habits for the age of AI. Next, learn to spot AI-written phishing and switch on the free AI defenses already on your phone.

Sources

This guide is general information, not legal or professional advice.