I was looking for a new show to watch while traveling, and I clicked on the Netflix show The Night Agent. The main character, Peter Sutherland, has to use code words to verify identity before trusting critical information over the phone. It seemed like spy drama fiction, the kind of thing that happens in high-stakes government operations, not real life.
Except now, cybersecurity experts say every family needs the same system. Not for espionage, but for protection against AI scams that can perfectly clone your voice in seconds.
In early 2024, a finance worker at a multinational company wired $25 million to scammers after a video call in which her CFO and colleagues urgently requested the transfer.
The problem was that none of them were real. Every person on that call was an AI-generated deepfake.
If you think AI scams still look like fake emails, it’s time to reconsider. Technology has advanced, making it difficult to trust what you see and hear.
And these scams are hitting families and small businesses everywhere.
Your Voice Isn’t Yours Anymore
Scammers only need three seconds of your voice, pulled from social media, a voicemail, or any public recording, to clone it using AI.
Then, they can make “you” say anything. The result can fool even those closest to you. Imagine a loved one gets a 2 a.m. call. It’s your voice, sounding panicked, asking for money. Would they stop to verify, or send it immediately?
The Family Angle Gets Personal
The scammers aren’t just calling the elderly anymore.
They study your social media to know when you’re traveling, when your kids are at school, and when family members have health issues. Then they strike at the most vulnerable moment with the most convincing story.
The emotional manipulation is what makes it work. When you hear someone you love in distress, your instinct is to help immediately, not to stop and verify. That’s exactly what scammers count on.
Small Businesses Are Getting Hit Too
Local businesses face a form of fraud known as CEO fraud. An employee receives an email from the owner asking them to process an urgent wire transfer or update vendor payment information. The email looks legitimate. Sometimes there’s even a voice call or video message to back it up.
Why it works: Scammers use LinkedIn, the company’s website, and public records. They know their vendors and processes, and then accurately impersonate leadership with AI.
The Defense That Actually Works
The good news: cybersecurity experts, including the National Cybersecurity Alliance and AARP, recommend a straightforward solution.
Establish a safe word.
Sit down with your family today and pick a word or phrase that only you will know—something random that wouldn’t come up in a normal conversation, and that a scammer scraping your social media couldn’t guess. The recipient of an urgent call could request that their family member say the pre-determined safe word to make sure they are not speaking with an AI impersonation.
What Makes You Vulnerable
Every video you post on social media. Every voicemail greeting. Every podcast appearance. That’s all the material AI can use to clone you.
For business owners, consider which information about your company, such as your vendors and processes, is public. Scammers examine your LinkedIn page and website.
The Bottom Line
You can’t always trust what you see or hear; verification is essential. That’s not paranoia. It’s reality.
Deepfake video technology can now operate in real time. According to BBC research, traditional detection methods, such as checking for unnatural movement or audio-syncing problems, are less reliable. The tools that help spot fake videos are becoming obsolete.
The safe word solution is simple yet effective. No matter how advanced AI becomes, it can’t tell what it hasn’t learned.
___
Photo, top, courtesy of Kampus Production / Pexels
