Our family safe word took eleven minutes to set up, including the argument about whether “meatball telescope” was too silly. (It wasn’t. Silly is memorable.) Those eleven minutes are now the single cheapest piece of security we own — no subscription, no app, no battery to charge — and it directly counters the scam category the FTC says cost Americans $3.5 billion in 2025: imposter calls. It’s also the same verify-before-you-act approach the FTC’s family emergency scam guidance recommends.
This guide is the tactical companion to our full AI voice cloning protection playbook. Here we’ll cover exactly how to choose a word, when to use it, word-for-word scripts for kids and grandparents, and the mistakes we’ve seen make safe words useless.
Why a safe word beats every other defense
Scammers can now clone a family member’s voice from a few seconds of social media audio. That means “it sounded exactly like her” is no longer evidence of anything. Caller ID can be spoofed. Photos can be generated. Even video is getting shaky.
What can’t be faked is a piece of information that exists only inside your family and has never been written down online. A safe word converts an impossible test (“does this voice sound real?”) into an easy one (“does this caller know the word?”). That’s the whole trick, and it’s enough.
Rules for choosing a safe word
Rule 1: Random beats meaningful
The instinct is to pick something sentimental — the old dog’s name, the street you grew up on, Grandpa’s nickname. Resist it. All of that is guessable, and much of it is sitting on Facebook right now. Scammers who research a family before calling (and the ones running higher-value scams do) will try exactly those words.
The best safe words are two unrelated everyday words: “orange ladder,” “velvet pancake,” “winter bicycle.” Nonsense pairs are easy to remember precisely because they’re absurd.
Rule 2: Everyone must be able to say it under stress
Test it on the youngest and oldest members of the family. If Grandma can’t recall it on the spot, or a seven-year-old can’t pronounce it, pick again. A safe word that requires calm recall fails at the exact moment it’s needed.
Rule 3: It travels by voice only
Share the word in person or on a phone call that you initiated. Never in a text, group chat, email, shared note, or family calendar. The moment it’s written in any account, it’s only as safe as that account’s password — and if you want to know how safe that is, see rule 4.
Rule 4: Protect the accounts around it anyway
A scammer inside a family member’s email or Instagram gets the contact list, the voice samples, and the personal details that make the con convincing. Locking down those accounts shrinks the attack before it starts. We moved our whole extended family onto 1Password Families for this reason — one afternoon of setup, and the “password is the dog’s name everywhere” problem disappeared.
Rule 5: Rotate on exposure, not on a schedule
You don’t need to change the word monthly. Change it when something happens: it got said in front of guests, a kid blurted it in a game chat, or — best case — it actually stopped a scam call, which means someone was probing your family and heard you ask for it.
When to use it (and when not to)
The trigger conditions should be mechanical, not judgment calls. In our house, the word is required whenever a call, text, or voice message involves any of these:
| Trigger | Examples |
|---|---|
| Money moving | Wire transfers, gift cards, Venmo/Zelle, crypto, bail, “lawyer retainers” |
| An emergency involving a family member | Accident, arrest, hospital, kidnapping, stranded abroad |
| Secrecy | “Don’t tell Dad,” “the case is confidential,” “you’ll get me in trouble” |
| Unusual instructions | Reading out a verification code, leaving school early, meeting a courier |
And two things the safe word is not for:
- It’s not a daily password. Don’t use it for pickup arrangements or routine calls. Every use in a non-emergency is a chance for it to leak, and overuse trains kids to say it to anyone who asks.
- It’s not a substitute for the callback. Even with the correct word, any request to move money still gets the hang-up-and-call-back-on-a-known-number treatment. The word filters out 99% of scams; the callback catches the rest.
Scripts: the exact words to use
Rehearsing feels awkward. Do it anyway — once, out loud, at dinner. People do under stress what they’ve practiced, and only that.
For grandparents receiving a “grandchild in trouble” call
“I hear you, sweetheart. Before anything else — what’s our family word?”
Then, no matter what happens:
“I’m going to hang up now and call you right back.”
Rehearse the hang-up specifically. Older adults were raised to consider hanging up rude, and scam scripts exploit that politeness ruthlessly. The family agreement — stated explicitly, ideally more than once — is: “We will never, ever be upset that you hung up and called back. Not even if it was really us.”
For kids and teens receiving a “parent” call
“What’s our word?” — and if the caller doesn’t have it: hang up, then call Mom or Dad directly from your own contacts.
Add one more rule for teens: the safe word also applies to texts and DMs asking for money, gift cards, or login codes, even from a sibling’s account. Accounts get hijacked; the word travels by voice, so a text can’t contain it — which means any money request by text gets a voice call before anything else.
For parents receiving a “kidnapping” or “accident” call
These are the hardest calls to think through, which is the point of scripting them in advance:
“I understand. I need to hear [child’s name] say our family word.”
While asking, text or call the child’s actual phone from another device, or call the school/friend they’re supposed to be with. Virtual kidnapping scams rely on you not attempting contact; in the overwhelming majority of cases, the “kidnapped” child answers their own phone, confused, from math class.
Mistakes that make safe words fail
We’ve heard from readers who set up a safe word and got burned anyway. It’s almost always one of these:
- The word was guessable. Pet names, maiden names, street names — all researchable. Random pairs only.
- It lived in a group chat. One compromised phone and the word belongs to the scammer, upgraded from defense to weapon (“See? I know the word.”) — the worst possible outcome. Voice only.
- Nobody rehearsed. A word nobody remembers under adrenaline is decoration. One dinner-table drill fixes this.
- It replaced the callback instead of adding to it. Layers, not substitutes. Word first, callback always.
- Asking felt rude, so nobody asked. This is the big one. The family norm has to be explicit: asking for the word is an act of love, and the correct response to being asked is “good job,” never offense.
- Only the grandparents got the talk. Teens are targeted too — sextortion follow-ups, fake “mom’s phone broke” texts, gaming-account cons. Everyone with a phone gets the word and the rules.
A 10-minute setup checklist
- Gather everyone — video call is fine for remote family, as long as you started the call.
- Pick two random words together. Veto anything connected to your family’s public life.
- State the four triggers: money, emergency, secrecy, weird instructions.
- State the two promises: asking is never rude; hanging up to call back is never rude.
- Run one 60-second drill: one person plays the scammer, someone asks for the word, hangs up, calls back.
- Repeat the word out loud once more. Done. Nothing gets written down.
While everyone’s assembled, it’s a decent moment for the broader five-minute family check: is two-factor on the main email accounts, are the kids’ profiles private, does anyone’s password still date from 2019? Our family security guides cover each of those in order of impact.
The bottom line
A safe word is the rare security measure that’s free, takes ten minutes, requires no technology, and directly defeats the fastest-growing scam aimed at families. Choose a random two-word phrase, share it by voice only, tie it to clear triggers, and rehearse it once. Then back it with the one tool that protects the accounts scammers mine for their material — we set our own parents up on 1Password Families, and between that and the safe word, the two most common paths into a family’s finances are closed.