AI Voice Cloning Fraud: Scammers Use Synthetic Voices to Drain Bank Accounts
A California mother lost thousands of dollars to an AI-powered voice cloning scam where fraudsters mimicked her daughter's voice in a fake kidnapping scenario. The FBI reports Americans lost over $893 million to AI-related fraud last year.
AI Voice Cloning Fraud: Scammers Use Synthetic Voices to Drain Bank Accounts
A mother in California reported losing thousands of dollars within a single month after receiving a phone call where she heard what she believed was her daughter in a state of panic. She now believes the call was part of a sophisticated fraud scheme created using artificial intelligence technology — one of the most widespread cases of so-called "voice cloning."
According to ERTnews, which cites CNN, this incident is part of a broader wave of fraud that exploits AI tools capable of producing highly convincing replicas of human voices from just a few minutes of audio material. Data from the U.S. Federal Bureau of Investigation (FBI) shows that Americans lost more than $893 million last year to AI-related scams, including attacks using cloned voices and AI-generated phishing emails.
Law enforcement and cybersecurity experts warn that these attacks can impersonate anyone — from family members and friends to colleagues or even service providers. Banks such as Britain's Starling and Australia's Commonwealth Bank have already alerted their customers about this type of fraud. Meanwhile, specialists note that the technology has advanced to a point where most people can no longer reliably distinguish a synthetic voice from a real one.
"For the average person, it's not fair to expect them to detect these kinds of scams," said Henry Ajder, an expert in AI-generated content. "Even I struggle. Most people do."
Voice cloning scams typically work by using a small audio sample from the target, often harvested from social media or previous phone calls. The perpetrators then use this material to create a digital copy of the voice and stage crisis scenarios — such as fake kidnappings or arrests — demanding immediate money transfers.
"There was no time to think," Philadelphia-based attorney Gary Schildhorn told CNN after being targeted by a scam that mimicked his son's voice. "It was simply: I need to act to help my son. He's in danger."
In more advanced cases, attackers may use text-to-speech tools or "voice skinning" techniques that allow real-time voice transformation during live conversations. Additional methods like caller ID spoofing can make calls appear to come from familiar numbers, further strengthening the deception.
When it comes to protection, experts say traditional signs of synthetic voices — such as pauses or unnatural inflections — are no longer reliable due to technological progress. Instead, they recommend recognizing general fraud indicators, such as time pressure, demands for secrecy, and urgent requests for money transfers.
In many cases, the most effective defense is verifying the caller's identity through another channel, such as a text message or a second call to a known number. Some specialists also suggest establishing a pre-arranged "code word" among family members that can be used to verify identity during emergency situations.
"If you suspect something is wrong, it's far better to have your child or friend tease you for thinking they were a robot than to rush to an ATM," Ajder noted.
Source: ERTnews
Source: ERT