SK Telecom said on the 1st that it has newly introduced the call security feature "AI voice phishing detection" to its artificial intelligence (AI) phone service "adot phone."
"AI voice phishing detection" analyzes the content of a call in real time and immediately notifies the user if the call is suspected to be a voice phishing attempt. SK Telecom said it expects the feature to protect users from increasingly sophisticated voice phishing crimes and provide a safer calling environment.
Whether a call is suspected of voice phishing is determined in real time during the call by an AI model that comprehensively analyzes various factors, such as the presence of suspicious keywords and conversation patterns. However, calls from numbers saved in the user's contacts or from official numbers registered in adot phone's "business contacts" are excluded from analysis.
Analysis results are classified into two levels, "suspicious" and "danger," depending on severity, and the user is immediately alerted during the call via a warning pop-up, alert tone, and vibration. If the user does not tap "Not voice phishing" on the warning pop-up, a "phishing detected" label is attached to the number after the call ends. The warning label can be viewed on screens such as "recents" and "search" within the adot phone.
If a user places a call to a number with a warning label, the adot phone displays a warning pop-up to confirm whether to proceed. If a call comes in from a suspicious number, the adot phone displays warning messages such as "Call detected as voice phishing" on the incoming screen to alert the user to the risk.
The feature processes the entire sequence—from analyzing call content to delivering warning alerts—on the device, based on On-device AI technology. The AI model analyzes the call content in real time on the device, continuously detecting signs of voice phishing. The company said this approach enables safe detection with no risk of information leaks, as call data does not pass through a server and there is no separate data storage or deletion process.
"AI voice phishing detection" is available through the adot phone app on the Android operating system (OS) and under the "phone" menu in the adot app on iOS. To use the feature, users must enable "AI voice phishing detection" in the "AI security" menu in settings.
The "AI security" menu consolidates key security features of the adot phone, such as "AI voice phishing detection" and "AI safe blocking," in one place. The menu also adds a "call interception detection" feature (available only on AOS) that detects attempts by malicious apps to change the number dialed by the user to another number and warns the user.
In addition, SK Telecom made it possible to check the contact information of relevant reporting agencies directly by entering keywords such as "voice phishing" and "smishing" on the "search" screen.
Cho Hyun-deok, who oversees the adot phone at SK Telecom, said, "To strengthen the security of the adot phone, we added 'AI message,' which detects and warns of suspected spam and phishing texts, in September, and we have now added the call security feature 'AI voice phishing detection.' We will continue to advance AI-based security features so customers can always enjoy a safe and comfortable calling experience."