1 in 3 People Got a Fake AI Voice Call Last Year

The Numbers

In June 2026, call-protection company Hiya released a global report on phone scam trends. The findings were striking.

Nearly 1 in 3 consumers worldwide reported receiving an AI deepfake phone call in the past year, a call where the voice mimicked someone familiar: a family member, a coworker, a friend.

In the U.S., the average person receives 10 unwanted calls per week. One in 12 Americans lost money to a phone scam, with an average loss of $682.

That's a real pattern of harm, spreading fast.

Trust in Phone Calls Is Breaking Down

The Hiya report found that 34% of consumers globally said their trust in phone calls declined over the past year. More than a third of people have grown skeptical of their own phone.

That erosion has consequences. People stop answering. They miss calls from doctors, schools, and family members in genuine emergencies. Scammers win twice: once when they steal money, and again when they make everyone afraid to pick up.

AI voice fraud doesn't just take money. It takes the simple, human reliability of a phone call.

How AI Voice Cloning Works

Voice cloning no longer requires sophisticated equipment or weeks of training data. Modern AI tools can replicate a person's voice from as little as 20 to 30 seconds of audio, pulled from a voicemail, a short video, or a social media clip.

Once cloned, that voice can say anything. It can call a parent in a panic, claim to be stranded, arrested, or hurt, and ask for money wired immediately, before anyone has time to think.

The voice sounds right. The urgency feels real. And the window to question any of it is deliberately narrow.

Why Your Ears Can't Save You

Human ability to detect synthetic audio is genuinely limited. Research on deepfake detection puts human accuracy at around 55 to 60 percent, barely better than a coin flip, and that's under controlled conditions without emotional pressure.

When the voice sounds like your child, your parent, or your sibling, the emotional weight overrides skepticism. That's not a failure of intelligence. It's a feature of being human, and scammers design their attacks around it.

Asking yourself "does this sound like them?" is no longer a reliable test.

The Grandparent Scam, Upgraded

The classic grandparent scam has a long history. A caller pretends to be a grandchild in trouble, asks for bail money or emergency funds, and pleads for secrecy. It has worked for years because it exploits love and urgency.

AI voice cloning made it worse. Now the caller doesn't just claim to be a grandchild. They sound like one.

Families across the country have wired thousands of dollars before thinking to call back on a known number. By then, the money is gone, the real family member is fine, and the confusion and shame that follows can linger long after the financial loss.

The problem isn't gullibility. The problem is that the tool used to verify identity, the voice itself, has been compromised.

Something Has to Change

If the voice can be faked, you need a different way to verify.

Calling back on a known number helps, but it takes time and requires the presence of mind to do it while scared. Watching for behavioral tells in the conversation can work, but it puts the burden on the person being targeted. What families need is something simple, shared, and impossible for a scammer to know, something that doesn't rely on sounding right.

Three Words

Trust Onion is a free app that gives families a shared set of three rotating codewords. The words change every few hours, calculated locally on each person's phone, with no server required. It works offline.

When a call comes in from someone claiming to be a family member, you ask one question: "What are the words?"

If they know them, it's really them. If they hesitate, deflect, or get it wrong, you have your answer.

AI can clone a voice in seconds. It cannot know three codewords that rotate on a private schedule and never touch a central server. The words aren't stored anywhere a scammer can reach. They're shared between people who trust each other, and nowhere else.

The voice might be convincing. The urgency might feel real. But "What are the words?" ends the uncertainty in three seconds.

One Question. Real Peace of Mind.

The Hiya report makes clear that AI voice scams are widespread, growing, and already costing American families hundreds of millions of dollars a year.

The solution doesn't require technical knowledge or expensive software. It requires agreeing on three words with the people you love, and knowing to ask for them when something feels off.

Simple things work. This one works really well.