Sungkyunkwan University research team develops voice security technology to detect spoofed audio

A research team at Sungkyunkwan University has created a voice security technology to counter growing threats from voice deepfake crimes, the university announced on Wednesday. Led by Professor Park Eun-il from the Department of Applied Artificial Intelligence, the team showcased how the technology can identify manipulated audio inserted into authentic voice recordings.

The innovative technology aims to find applications in various security-related fields, such as preventing voice phishing, verifying identities through voice-based authentication for financial services, and analyzing digital evidence. To achieve this, the researchers employed artificial intelligence (AI) to examine minute alterations in audio signals and speech patterns that are challenging for the human ear to detect. The AI system precisely locates the areas of manipulated speech down to one-second intervals.

Two papers authored by the team were accepted for presentation at the Interspeech 2026 conference, an annual gathering of the International Speech Communication Association on speech science and technology, taking place from September 27 to October 1 in Sydney, Australia.