.png?w=1024)
A familiar voice is no longer proof that a call is real.
A California mother wired $5,400 after scammers allegedly made her believe her daughter had been kidnapped, according to ABC7. The case reflects a growing fraud risk as criminals use AI voice-cloning tools to imitate family members, create panic, and pressure victims into sending money before they can verify what is happening.
For consumers and security teams alike, the lesson is blunt: trust the process, not the voice.
How scammers are using AI voices
Fraudsters typically begin by gathering short audio clips from public sources such as social media videos, podcasts, interviews, or even voicemail recordings. These clips are then fed into AI software that studies tone, rhythm, and accent to recreate a synthetic version of the target’s voice.
Once the clone is created, scammers use it to make urgent phone calls, often posing as family members in distress. In one reported case, a victim described hearing what sounded like his son pleading for help and asking for money for bail. The emotional pressure, combined with a sense of urgency, is designed to push victims into acting quickly without verification.
Security experts say newer techniques can even enable real-time conversations using manipulated or AI-generated speech, making the deception even harder to detect.
Real victims, real losses
In a case reported by ABC7, California resident Deborah Del Mastro said she lost $5,400 after receiving a call claiming her daughter had been kidnapped by a cartel.
She told reporters she heard what she believed was her daughter’s voice in panic, saying: “I love you, mom, I’m so sorry, I’m so scared.” The caller issued strict instructions over several hours, including repeated warnings such as “don’t speak” and “it has to happen now,” according to Del Mastro’s account.
She later discovered her daughter was safe and at work, confirming the entire incident was a scam. Authorities and investigators say these cases are part of a trend of impersonation fraud increasingly powered by AI tools.
Why these scams are harder to spot
Traditionally, people were advised to listen for robotic tones or unnatural pauses as signs of fake audio. But experts say those indicators are rapidly disappearing.
AI-generated voices are now so realistic that most people cannot reliably tell the difference between real and fake audio. Scammers also use caller ID spoofing to make it appear as though calls are coming from trusted contacts, further increasing credibility.
According to experts, the real danger lies not in how the voice sounds, but in how the message makes the victim feel: urgency, fear, and pressure to act immediately.
The growing financial toll
The scale of AI-enabled fraud is expanding quickly. The FBI reported that Americans lost $893,346,472 last year alone to AI-related scams, including voice cloning, phishing schemes, and impersonation attacks.
Criminals frequently request payments through methods that are difficult to trace or reverse, such as wire transfers, cryptocurrency, or gift cards. Security analysts say the combination of emotional manipulation and advanced AI tools is making these scams more effective than traditional phone fraud.
How to protect yourself
Experts across law enforcement and cybersecurity agencies emphasize that verification is now more important than voice recognition.
Common advice includes hanging up and calling the person back using a trusted number, rather than the one that contacted you. Families are also encouraged to create a code word known only to trusted members, which can be used to confirm identity in emergencies.
Authorities also warn people to be cautious of any unexpected requests involving money, secrecy, or urgency, three of the strongest warning signs of a scam.
Also read: AI-driven social engineering topped ISACA’s 2026 cyber-risk concerns as deepfakes, voice cloning, and other AI-enabled attacks tested security teams.