Are You Safe From AI Spoofing? Deepfake Voice Sparks Concerns Around The World
Updated: Dec 19, 2023
Criminals around the world are taking advantage of modern advancements in artificial intelligence (AI) technology to deceive and steal from potential victims.
According to CBS News, it's now so easy to spoof a person's voice using AI that it can be done in just five minutes. In an episode of the network's series '60 Minutes', Rachel Tobac, an ethical hacker and CEO of the data protection group 'Social Proof Security', tries to spoof the voice of a correspondent, which she is able to do so using existing clips of the show that aired on television.
While it'll, of course, prove easier to spoof the voice of a public figure who can be heard in many recordings, Tobac claims that no one's voice is 100% safe. "Oftentimes attackers will go after people, they don't even know who these people are, but they just know this person has a relationship to this other person," said Tobac to 60 Minutes.
What perhaps is more concerning, as Tobac warns, is that such schemes are convincing enough to fool anyone, from seniors to even tech-savvy millennials. A Federal Bureau of Investigation (FBI) report found that Americans lost more than US$10 billion last year to online scams and digital fraud, with seniors losing the most to scammers and those in their 30s filing the most complaints.
Susan Monahan, an 81-year-old tax preparer with an MBA, told 60 Minutes that she fell victim to a grandparent scam and had US$9,000 stolen from her. She recalled receiving a call from a young man who was posing as her grandson, saying he was in trouble and needed help to avoid being jailed. A man pretending to be an attorney then got on the line and walked Monahan through what she had to do. Monahan, not knowing she was being scammed, followed the instructions, which involved withdrawing cash and handing it to a courier who would show up at her home. It was only after she handed the money that she realised she was scammed.
"Most hacks start with some sort of social engineering or human-based hacking because that's the easiest thing to do," Tobac added.
Tobac further warns that hackers no longer need to infiltrate computers through a back door to steal money or launch schemes. Up to 95% of hacks today supposedly happen after a user clicks on a malicious text or link which automatically steals their personal information.
That's why, statistically, Americans are more likely to become victims of theft online than a physical break-in.
Cases of online and AI-related theft aren't limited to the U.S. as there have also been a number of reported incidents in China.
According to The Straits Times, a fraudster managed to impersonate a man's friend during a video call and convince him to part with 4.3 million yuan, which is equivalent to about S$822,000. The fraudster reportedly used AI-powered "deep fake" technology to deceive the victim, as reported by police in the city of Baotou within the region of Inner Mongolia. The victim transferred the money thinking that his friend needed it to make a deposit during a bidding process and only realised he had been tricked after he found out his actual friend hadn't a clue about the transfer.
The police said some of the funds have been recovered and they're working to trace the rest. The incident went viral on Chinese social media, with many expressing concerns over the potential of AI tech to be used for malicious purposes.
"This shows that photos, voices and videos all can be utilised by scammers," one user said. "Can information security rules keep up with these people’s techniques?"
Criminals around the world are taking advantage of modern advancements in AI technology to deceive and steal from potential victims.
CBS News hired an ethical hacker to try and spoof the voice of one of its correspondents and she was able to do it in five minutes with the help of AI.
She warns that hackers no longer need to infiltrate computers through a back door to steal money or launch schemes.
In China, a fraudster managed to impersonate a man's friend using "deep fake" technology and convince him to part with 4.3 million yuan.