A scary call. A frantic 911 report. Police racing to stop what they thought was a kidnapping – only to learn that it was all a hoax.
Such was the case recently in Lawrence, Kan., where a woman picked up her voicemail to find it hijacked by a voice eerily like her mother’s claiming to be in trouble.
The voice was AI-generated, actually pretty fake. And all of a sudden, it wasn’t the plot of a crime novel – it was real life.
The voice on the other end “sounded exactly like her mom,” police say, matching tone, inflection, even a heightened emotional state.
The whole thing feels like scammers took some public audio (perhaps from social media or voicemail greetings), fed it through some voice-cloning AI, and watched the world burn.
So the woman dialed 911; police traced the number and pulled over a car - only to find: no kidnapping. Only a virtual threat meant to deceive human senses.
It’s not the first time something like this has occurred. With just a snippet of audio, today’s artificial intelligence can generate the dulcet tones of Walter Cronkite or, say, Barack Obama – regardless of whether the former president has said anything like what you’re hearing..segments using deep fakes to manipulate people’s actions in new and convincing ways.
One recent report by a security firm found that about 70 percent of the time, people had trouble distinguishing a cloned voice from the real thing.
And this isn’t just about one-off pranks and petty scams. Scammers are deploying these tools to parrot public officials, dupe victims into wiring them vast sums, or impersonate friends and family members in emotionally charged situations.
The upshot: a new kind of fraud that’s more difficult to notice – and easier to perpetrate – than any in recent memory.
The tragedy is that trust so easily becomes a weapon. When your ear – and your emotional response – buys what they hear, even the basest gut-checks can vanish. Victims often don’t realize the call was a sham until it’s far too late.
So what can you do if you receive a call that feels “too real”? Experts propose small, but crucial safety nets: pre-established “family safe word,” check by calling back your loved ones on a known number and not the one that called you, or ask questions only real person would know.
OK, so it’s old-school phone check, but in the era of AI that can reproduce tone, laughter even sadness – it could be just the ticket for keeping you safe.
The Lawrence case in particular is a wake-up call. As AI learns to mimic our voices, scams just got much, much worse.
It’s not just about fake emails and clicking on phishing links, anymore – now it’s listening to your mother’s voice on the phone, and wanting with every atom of your being to believe that something horrible has not taken place.
That’s chilling. And it means that all of us have to stay a few steps ahead – with skepticism, verification and a happy dose of disbelief.
