Four major new methods including deep voice and AI phishing... There is a need to prepare password protection among family members.

Four major new methods including deep voice and AI phishing... There is a need to prepare password protection among family members.

(Seoul = Yonhap News) Reporter Hayoung Kwon = Ahead of the Chuseok holiday, a new type of holiday fraud that exploits artificial intelligence (AI) technology has been revealed.

East Security, a security specialist subsidiary of EastSoft [047560], announced four major types and urged caution: deep voice and deepfake video call fraud, AI-generated customized phishing with no grammatical errors, 'all-in-one account takeover' that takes over banking, messengers, and social networking services (SNS) at once, and ultra-precise impersonation combined with personal information leakage incidents.

What these types have in common is that they occur the moment the victim believes they have “seen it themselves.”

Recently, technology that synthesizes the voices and faces of acquaintances in real time with just a few seconds of voice samples is being abused for crime. It is necessary to set up a separate verification method in advance, such as a password that only family members can know, or to reconfirm by calling the number you already know when receiving an urgent request for money.

With the development of generative AI, grammatical errors that were a feature of past phishing texts have disappeared.

If a remote control app is installed through a text message disguised as a courier or fine, not only financial information but also messenger and SNS accounts can be stolen, so installation of apps from unknown sources must be blocked at the source.

An East Security official said, “The existing fraud detection method that relied on human senses has now become the target of AI attack,” and asked, “If you discover suspicious circumstances, please immediately report it to the National Police Agency or the Financial Supervisory Service.”