
A recent case illustrates how AI-powered apps are helping Koreans avoid voice phishing scams. When a caller claiming to be a police investigator asked Mr. A to click a URL regarding an alleged sports gambling site issue, his smartphone app triggered a phishing alert, prompting him to hang up immediately and report the number to authorities.
In another instance, Ms. B received a call from someone impersonating a corporate marketing team, offering free products as part of a review program. When the caller requested she add them on KakaoTalk to send product information, her smartphone's phishing detection app issued a warning, allowing her to end the call before sharing any personal information.
With the Lunar New Year holiday approaching, warnings about voice phishing have intensified. Fraud cases targeting elderly victims typically surge around major holidays, with criminals posing as delivery services, family members, or government subsidy programs.
According to the Ministry of Science and ICT, Samsung Electronics and Korea's three major telecom carriers—SK Telecom, KT, and LG Uplus—now offer AI-based voice phishing detection services. These services utilize on-device AI technology to analyze calls in real time and alert users to potential threats.
The detection features are available through Samsung's Phone app, SK Telecom's A. Phone app, KT's WhoWho app, and LG Uplus's ixi-O app. All services analyze incoming calls from unknown numbers in real time, classifying risk levels as either "suspicious" or "warning," and notify users through pop-ups, alert sounds, or vibrations during calls.

Samsung's Galaxy smartphones include a built-in "Suspicious Voice Phishing Call Alert" feature. When risky conversation patterns are detected, the screen displays either "Suspected voice phishing" or "Warning: Voice phishing detected" for higher-risk situations. The feature is available on Galaxy devices running One UI 8.0 or higher, regardless of carrier, but not on other Android phones or iPhones.
SK Telecom's A. Phone app provides detection through AI security settings, available on Android devices with the pre-installed app or on iPhones (XS, XR, SE 2nd generation or later running iOS 17+) for SK Telecom subscribers only.
KT's WhoWho app combines real-time conversation analysis with speaker recognition and deepfake voice detection technology. The AI, trained on criminal scenarios, compares voices against reported fraudsters' voiceprints. The service works on Android phones running OS 9 or higher (Galaxy S8 and later), regardless of carrier, but is not available on iPhones.

LG Uplus's ixi-O app offers "Anti-Deepfake" technology to identify manipulated voices and can detect matches with reported criminals' voices. The service is available to LG Uplus subscribers on Android devices running OS 14 or higher (Galaxy S21, Z Fold3/Flip3 or later) or iPhones running iOS 17 or higher.
These services process call data entirely within the smartphone using on-device AI, minimizing data breach risks and privacy concerns by avoiding transmission to external servers.
"Voice phishing has recently evolved to use AI to maximize psychological pressure on victims," the Ministry of Science and ICT said. "With crimes likely to increase around the Lunar New Year holiday, we urge people to actively use their smartphones' voice phishing detection features."
Experts emphasize that simply checking and configuring detection settings on elderly parents' smartphones can significantly reduce the risk of fraud.
