The best defense against high-tech AI scams is low-tech skepticism, expert warns
AI‑Generated Voice Scams Target American Families, Experts Urge Human Skepticism
In a startling illustration of how quickly artificial‑intelligence tools can be weaponized for fraud, a California mother transferred more than $5,000 to scammers who had cloned her daughter’s voice to stage a kidnapping hoax. The incident, highlighted by cybersecurity specialist Morgan Wright on the CyberGuy Report podcast, underscores a growing threat to everyday citizens who may be tricked into acting on lifelike audio or video that seems to come from a loved one.
The fraud was carried out by criminals who harvested a brief clip of the child’s voice from social media. Using generative voice‑cloning software, the scammers recreated the girl’s voice so convincingly that the mother believed she was hearing a real distress call. “It’s my daughter’s voice,” Del Mastro said. “And she’s having a panic attack.” With the illusion of imminent danger in place, the mother transferred more than $5,000 to the scammers before realizing she had been duped.
Wright, who has trained federal, state, and local law‑enforcement teams—including the FBI and the National Security Agency—points to a broader trend of “hyper‑realistic” scams that rely on psychological manipulation rather than sophisticated detection algorithms. “When you can fool a mother, that’s how good these things are getting,” he said. “What they do is use manipulation, influence, and deceit. The more urgent I can make it sound, the more you’re going to lower your guard and take action.”
The technology behind these scams is part of a rapid evolution in AI that can eliminate language barriers, produce convincing phishing texts, and generate audio and video that mimic real people. Even with the rollout of new security features, criminals are constantly finding ways around software‑based detection because AI models—such as those from xAI, Anthropic, OpenAI, and Gemini—undergo frequent changes. “We hear about all the different models … they change models constantly,” Wright noted. “So when the model changes, how effective is detection software? We don’t know because this stuff is moving so fast.”
In the United States, most scams are launched by individual cybercriminals rather than state actors, according to Wright. He cautions that every scam relies on basic psychological triggers: urgency, fear, and a sense of personal connection. His key recommendation to the public is a simple pause. “We’re not storming the beaches at Normandy; you’re not going to get arrested,” he said. “No law‑enforcement agency sends you an unsolicited text saying you’re about to get arrested.”
Wright’s guidance echoes a timeless warning he cited from the 1860s: “Believe nothing that you hear and only one half that you see,” referencing Edgar Allan Poe. “That advice is as good now as it was 160 years ago.” In a world where AI can instantly clone a voice from a few seconds of audio, the age‑old caution that human skepticism remains the best line of defense against fraud has never been more relevant.