Researchers at Abertay University have raised concerns that advances in AI voice technology could be exploited by scammers to target unsuspecting victims by using regional accents.

A new study, led by Dr Neil Kirk of the Department of Sociological and Psychological Sciences, highlights how realistic AI-generated voices have become, with listeners assuming the voices must be real, particularly when mimicking regional dialects like Dundonian Scots.

The research has raised worries among experts that scammers could use AI voice-generating tools to target specific areas of the country, adopting regional dialects to appear more human and deceive the unwary.

Published in the journal Computers in Human Behaviour: Artificial Humans, the research involved the creation of six common phrases in both Scottish Standard English and Dundonian dialect, for example, “The dog had five puppies” and “The doag had fev puppies.”  

The recordings were then processed using AI software, and when played to 100 Scottish adults and 100 people from Southern England, most listeners struggled to distinguish either version from human voices. 

Although Scots were slightly better at recognising the fake voices than English participants, many AI versions still succeeded in deceiving them.

“Scottish people are not often accustomed to technology understanding their way of speaking, let alone sounding like us,” said Dr Kirk.

“It’s not surprising that Scottish participants in this study were more likely to perceive voices speaking in Dundonian Scots as human, even when those voices were actually AI-generated.

“My research shows that the public often underestimates just how sophisticated this technology has become, while simultaneously overestimating their ability to detect a fake voice. As AI voices become more advanced, it’s getting harder and harder for most people to tell them apart from real human speech.”

Recommended

• Report: More Than 50% of Fraud Is Now Driven by AI
• Microsoft Thwarted $4BN in Fraud
• How Worried Are Brits About AI-Fuelled Phishing?

Concerns over the capabilities of AI tools to generate deepfake voices are already on the rise. 

A recent report from fraud prevention firm Feedzai 44% of financial professionals report they have seen deepfakes used in fraudulent schemes, with 60% recognising voice cloning in particular as a major concern.

Added to that, a YouGov poll taken in April found that nearly a quarter (24%) of financial phishing scams are now carried out via phone calls as part of criminals’ voice-phishing, or ‘vishing’, campaigns, compared to just 14% initiated through email. 

These AI-powered deepfake scams can hit the most vulnerable particularly hard. According to a recent report from Police Scotland, an elderly victim from Biggar was defrauded of around £2.14 million in a vishing and safe account scam, with their life savings being funnelled into untraceable gold bullion.