The Safety Brief
Plain-English safety guidance for families
ADVERTORIAL — We may be compensated when you sign up through links on this page. Details

Family Safety Alert

The 2 A.M. Phone Call Every Family With Aging Parents Needs to Prepare For

The FBI is warning that criminals are now using artificial intelligence to copy the voices of children and grandchildren — and calling the people who love them most. Three minutes of preparation stops the most common version of this scam cold.

A worried woman in her seventies sits on the edge of her bed at night, holding a phone to her ear
The calls come late, when judgment is groggy and fear is loud. The voice on the line sounds exactly like family. (Illustration of the scam pattern described by the FBI.)

Imagine the phone ringing at two in the morning.

The voice on the line is your grandson's. Not similar — his. The same laugh-roughened edge, the same way he says "Grandma, it's me." Except he isn't laughing. He says he's been in an accident. Or arrested. Or worse. There's crying. A second voice — a "lawyer," a "police officer" — takes the phone and explains, calmly, that this can all be handled quietly tonight: a few thousand dollars, by wire transfer or gift cards, and please, don't call his parents. He begged them not to wake anyone.

Almost nobody hangs up on that call. Because it isn't a stranger's voice. It's a voice you've known since the day he was born.

It's not him. And it only took three seconds of audio.

This month, the FBI issued a warning about exactly this scheme: criminals using AI "voice cloning" tools to mimic loved ones in distress. The technology needs only a few seconds of real audio to build a convincing copy of anyone's voice — and that audio is already public. A birthday video on Facebook. A graduation clip. A voicemail greeting. A school fundraiser livestream.

A family films a laughing grandfather with a smartphone at a backyard birthday party
The raw material is already online: a few seconds of a birthday video is all the technology needs to copy a voice.
$2.3 billion lost by older Americans to AI voice-cloning scams in 2026 so far, according to FBI figures reported in May — with adults over 60 absorbing 43% of all AI-fraud losses.

The demands typically run between $2,500 and $15,000 — paid in gift cards, wire transfers, or Bitcoin, all nearly impossible to recover. And the targets are not careless people. Retired teachers. Engineers. Nurses. The scam doesn't work because victims are foolish. It works because the call hijacks the two things that make someone a good parent or grandparent in the first place: love, and the instinct to act fast when family is in danger.

"Sharp, careful people lose money to this every night. Being smart is not a defense — having a plan is."

The one-word defense law enforcement recommends

Here is the good news: the most effective defense against voice-cloning scams costs nothing and takes one family conversation. Law enforcement and cybersecurity experts recommend every family agree on a safe word — a single word or phrase that only your family knows, and that any caller claiming to be family must say before you believe an emergency is real.

We've prepared a free, printable Family Safe Word Agreement — a one-page sheet families fill in together and keep by the phone. It takes about three minutes. You can get it at the bottom of this article.

A printed family agreement sheet held to a refrigerator door by a magnet
Three minutes, one conversation, one sheet by the phone — the defense that costs nothing.

The step most families miss

But there's an uncomfortable question hiding inside the safe-word advice, and it's the part most articles skip:

How did the scammers know who your grandson was — and your phone number — in the first place?

The answer is that they didn't guess. Data-broker websites openly sell profiles that connect your name, age, phone number, home address — and your relatives' names — assembled from public records and old data breaches. A criminal pairs that family tree with a voice sample from social media, and the 2 A.M. call writes itself. That's why the FBI tracked $893 million in AI-fraud losses last year, the first year it counted this category at all — and why this year is on pace to be far worse.

You cannot remove yourself from those databases once — there are hundreds, and they repopulate. What families increasingly use instead is a protection service that does it continuously: scrubbing your information from data-broker sites, screening scam calls, monitoring accounts and identity use around the clock, and backing it with fraud insurance and a 24/7 resolution team for the worst case. It's the same category of service banks recommend for identity theft — extended to the way scams actually happen now.

Step 1: Find out what scammers can already see about your family

Our partner offers a free exposure scan — it checks in seconds whether your name, phone number, and family connections are listed on data-broker sites and past breaches, then shows exactly what it found.

Run My Free Scan →

Free to check. No card details needed for the scan. Plans with 24/7 fraud monitoring and $1M identity-theft insurance start around $9/month if you choose to protect your household.

Step 2: Set your family's safe word tonight

Download the free Family Safe Word Agreement — the printable one-page sheet to fill in with your family and keep by the phone.

Get the Free Safe Word Sheet →

Do both. Tonight.

The families who lose $15,000 to a cloned voice and the families who calmly hang up are not different kinds of people. They're the same people, separated by one conversation they did or didn't have, and one check they did or didn't run. Three minutes for the safe word. Thirty seconds for the scan. Do it while it's on your mind — the calls are going out every night.