NTT East and NABLAS have launched a consortium initiative, “Development and Demonstration of Telephone Voice Fake Detection and Comprehensive Countermeasures against False and Misinformation on the Internet,” following the Ministry of Internal Affairs and Communications’ (MIC) adoption of the project.

Specifically, the two companies will develop highly accurate telephone voice fake detection technology compatible with the latest voice synthesis technology. The technology will maintain stable detection accuracy even in telephone voice environments where voice format and quality change.

The developed telephone voice fake detection technology will also be incorporated into a telephone app, demonstrating its operation on an actual telephone service. Voices synthesized by generative AI will be streamed on the telephone service, demonstrating the ability of the telephone app to detect fake voices under various conditions, including voice format and call patterns.

The widespread use of generative AI has made it easy to generate highly accurate images, video, and audio. Meanwhile, the misuse of this technology to spread fake and misinformation and commit specialized frauds has become a serious social problem both in Japan and overseas.
Spoofed phone calls in particular are more sophisticated and realistic than traditional fraud methods, and there has been a sharp increase in cases where victims are deceived without realizing the calls are fake.
Fake images and videos spread through social media and video sharing sites pose a risk of leading to incorrect judgments and confusion in fields such as politics, economics, and disaster response.
In response to this situation, the two companies will work to develop and implement technology that can detect fake and misinformation and protect the reliability of information.

※Translating Japanese articles into English with AI