Overview NABLAS Inc., an AI research institute (Headquarters: Hongo, Bunkyo-ku, Tokyo; Representative Director and Director: Kotaro Nakayama), has announced the completion of the development and demonstration of a 'telephone voice fake detection system' and 'disinformation countermeasure' technology. This work was conducted as part of the Ministry of Internal Affairs and Communications' 'Technology Development and Demonstration Project for Countermeasures against Disinformation (FY2025)', in a consortium with NTT East Corporation. The results are published on the Ministry's website.

Demonstration Details In this demonstration, the team constructed a voice dataset through actual telecommunications networks to verify a system that detects AI-generated audio sequentially and with low latency during calls. Furthermore, in collaboration with Ina City, Nagano Prefecture, they organized operational issues in local government settings and identified challenges for practical implementation.

Background of Development The rapid spread of generative AI has made high-precision audio generation easy, leading to serious social problems worldwide involving 'impersonation calls' and investment scams. To combat this, the social implementation of technology to protect the reliability of information is essential.

Future Outlook and Partner Recruitment Moving forward, the company aims to build a more robust detection system by implementing models adaptable to all environments and combining multiple detection methods. To accelerate the accuracy improvement and social implementation of voice fake detection technology, NABLAS is seeking joint demonstration partners, including local governments, prefectural police, financial institutions, telecommunications carriers, and companies providing online conference systems.

FACT BOX

  • Source: PR TIMES
  • Category: News