Juan David Gutiérrez, a professor and UN advisor, is warning the public about the rise of identity theft through AI voice cloning.

These scams represent a growing threat to digital security because they leverage highly realistic audio to deceive victims into sending money under false pretenses.

Speaking with Caracol Televisión, Gutiérrez said the mechanics of how AI-driven impersonation operates [1]. He said that criminals use these tools to mimic the voices of family members or officials to create a sense of urgency and trust. This tactic is designed to bypass the natural skepticism people have toward unknown callers.

One victim in the U.S. described a scenario where a caller claimed her daughter had been in an accident and required $15,000 immediately [3]. The high degree of vocal accuracy in these deepfakes makes it difficult for recipients to distinguish between a real person and a synthetic clone.

To combat these fraud attempts, Gutiérrez said measures for citizens to reduce their risk [1]. He said the importance of establishing "safe words," or verification protocols, within families to ensure the identity of the caller during emergency requests.

Andrés Vautrín, another expert on the subject, said that while technology is designed to help people, criminals are also utilizing it for illicit gains [4]. The ability to clone a voice from a short audio clip—often harvested from social media—has lowered the barrier for entry for cybercriminals.

Experts recommend that individuals avoid sharing excessive personal audio or video on public platforms where it can be easily scraped by AI software. When receiving a suspicious call, the safest course of action is to hang up and contact the person directly through a known, trusted phone number [1].

“The technology comes to help us, but criminals also use it.”

The shift toward generative AI audio signifies a move from traditional phishing—which relied on text and email—to social engineering that exploits emotional bonds. As voice synthesis becomes indistinguishable from human speech, the burden of verification shifts from the technology to the user, necessitating a cultural change in how people verify identity during remote communications.