Increasingly Sophisticated AI Voice Scams Demand Vigilance
Risks Rise with Emergence of AI Skillful in Mimicking Regional Accents

A recent study has found that people tend to be more easily persuaded by voices with a timbre similar to their own. As artificial intelligence (AI) technology evolves and becomes increasingly capable of mimicking a wide array of voices with great sophistication, there is a growing need to be cautious to avoid falling victim to scams that take advantage of these advances.


Professor Naykyung Hyun (Kimberly Hyun) of the University of Cincinnati and her research team presented these findings in a paper published in the Journal of Marketing Research in March of this year. The authors explained that, during investment pitches, the more similar the timbre of the presenter's voice is to that of the listener, the more persuasive the pitch tends to be. This relationship was verified using measurable outcomes, such as the amount of investment funds attracted.

"Familiar Voices Boost Trust... Easier Persuasion When Timbre Matches Your Own" View original image

The research team reached these conclusions by analyzing 14 seasons of the investment reality TV show "Shark Tank" aired by ABC in the United States. Professor Hyun explained to The Wall Street Journal (WSJ) that people tend to perceive those whose voices resemble their own as more competent, moral, and trustworthy.


The WSJ also introduced research suggesting that people are more likely to trust AI when it uses a regional accent. Neil Kirk, a professor at Abertay University in Dundee, Scotland, published a study in the Journal of Cybersecurity in January of this year. He analyzed what happened when Scottish people listened to both human and AI-generated voices delivering messages in both standard Scottish English and the local Dundee dialect.


Most of the Scottish participants in the experiment believed that the AI-generated voices were actually human. The rate of confusion was even higher when the AI used the Dundee dialect. Professor Kirk told the WSJ, "The technology has become so advanced that people naturally assume they are listening to a real human voice."


The WSJ noted that with the vast array of voice data available via social networking services (SNS) and other platforms, AI can now closely mimic local accents or even the voices of specific individuals, and scammers are exploiting these capabilities. As AI-powered scams become increasingly sophisticated, the WSJ cautioned that when receiving a phone call, it is important not to respond immediately but to critically analyze and verify the content of the call. It also recommended not drawing hasty conclusions about the caller’s identity based solely on their voice, but instead hanging up and contacting the individual via a separate, verified channel for confirmation.



If you receive a call from someone claiming to be a bank employee, you should immediately hang up and contact the bank directly using the official phone number printed on your transaction statement or bank card. Even if you get a call from someone impersonating a relative or family member and urgently requesting money, you should confirm with the person in question or another family member before taking any action. Regardless of how convincing the voice may sound, the WSJ stressed that you should never provide your password or any other sensitive authentication information to anyone who has called you first.


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing