KT Develops AI Service to Detect Voice Phishing Voices and Provide Phone Numbers

Voice Phishing Detection and Suspicious Number Alerts
Rapid Detection of Contexts Related to Financial Fraud
Customer Warning and Caution When Receiving Calls from Suspicious Numbers

KT announced on the 8th that it has developed a comprehensive voice phishing detection technology using artificial intelligence (AI) called the 'Voice Phishing Detection AI Agent' and the 'AI Voice Phishing Suspicious Number Alert Service' to protect customers.

The Voice Phishing Detection AI Agent is a technology that converts call audio into text in real time and quickly detects contexts related to financial fraud through AI. If suspicious signs of voice phishing are detected during the call, it immediately notifies the customer. This technology comprehensively verifies various risk situations such as ID cards, passwords, and monetary demands. From the perspective of personal information protection, it operates using an on-device method that does not send analysis data externally, which is an advantage in preventing information leakage. The National Forensic Service provided KT with anonymized voice phishing script data to develop the Voice Phishing Detection AI Agent.

The Voice Phishing Detection AI Agent is currently in beta service for KT internal experts. It will be officially launched this month by being integrated into the spam blocking application 'WhoWho.' With this, WhoWho will be reborn as an all-in-one service capable of detecting not only spam and smishing but also voice phishing.

The AI Voice Phishing Suspicious Number Alert Service is a technology that pre-detects phone numbers suspected of voice phishing using an AI model and notifies customers in real time. This technology learns the characteristics of phone numbers involved in voice phishing provided by the National Police Agency to detect suspicious numbers. If a call comes from a suspicious number, a warning video plays with a cautionary message before the customer answers. If the customer has already received the call, they are notified once more through a guidance call. The same method is used to guide customers when they make calls to suspicious numbers.

The AI Voice Phishing Suspicious Number Alert Service is currently being tested with employees. It is preparing to be released as an additional service that any KT customer can use in the future.

Yoon Kyung-ah, Executive Director of KT AI Tech Lab, stated, "We will do our best to reduce voice phishing damage, including developing speaker recognition technology that can detect even 'impersonation of acquaintances' using the voices of voice phishing criminals and deep voices."