"New Research Combats Burgeoning Threat of Deepfake Audio"

A new study by three UC Berkeley School of Information students and alums aims to make it easier to determine the authenticity of an audio clip as deepfakes and doctored audio have become more common. Deepfakes are a type of media, including images, audio, and videos, manipulated or created using Artificial Intelligence (AI). In addition to spreading false information, deepfakes can decrease the effectiveness of security systems. The team explored various techniques for distinguishing a real voice from a cloned one made to impersonate a specific person. According to Professor Hany Farid, the team has done important work in outlining different ideas for detecting the threat of deepfake audio. This article continues to discuss the new research on combating deepfake audio.

UC Berkeley School of Information reports "New Research Combats Burgeoning Threat of Deepfake Audio"

Submitted by grigby1