SK Shieldus senior official Lee Ho-suk delivers a speech at an industry seminar designed to deal with various phishing scams in Seoul on Tuesday. Photo courtesy of SK Shieldus
SEOUL, June 22 (UPI) — Artificial intelligence-based voice cloning technology is suspected of being used in phishing scams, according to South Korean security firm SK Shieldus.
During an industry seminar, the company, an affiliate of South Korea’s top mobile carrier SK Telecom, said generative AI programs like OpenAI’s ChatGPT and Google‘s Bard tools could already be being used in more sophisticated phishing attacks.
Advertising
“With just 5 seconds of a child’s dialogue file, the scammer can clone his or her voice,” SK Shieldus senior official Lee Ho-seok said at the Seoul meeting.
“Recently, there was a phone scam where a cybercriminal used a specific child’s cloned voice to demand ransom from parents,” he said.
Phone scams, or phishing, have been a huge problem in South Korea, with the perpetrators often targeting the elderly, who succumb to calls from scammers pretending to be family members or close friends.
Afraid the crisis is real, these victims end up sending money as requested.
The monetary damage has been on a steady rise in Korea. It was estimated to be $310 million in 2018, $540 million in 2020 and $600 million in 2021.
SK Shieldus expects AI-driven phishing scams to make inroads via other communication channels like instant messaging, email and websites.
Similar cybercrimes are raging in other countries as well, according to U.S.-based cloud security company Zscaler.
Its 2022 annual report noted that the five most targeted countries for phishing were the United States, the United Kingdom, the Netherlands, Russia and Canada.
“To fight back AI-driven phishing attacks, which can convincingly mimic legitimate communications, we also need generative AI-based technologies,” Lee said.
