Attack on Titan voice actor Yuki Kaji releases official voice synthesizer to counter people feeding his voice to AI without permission

Japanese voice actor Yuki Kaji, who played Eren Jaeger in Attack on Titan, released his official voice synthesis software on November 21.  

The synthesis software features Kaji’s alter ego Soyogi Soyogi, designed especially for this project. Two versions have been released, one that reproduces Kaji’s speaking voice and the other which replicates his singing. Both voices are available on CeVIO AI, VoiSona and VoiSona Talk. 

Yuki Kaji first announced he was working on this back in September 2023, as a project to mark 20 years of working as a voice actor. He crowdfunded the project through Campfire, with an initial goal of 10 million yen, but ended up raising over 34.1 million yen, three times the amount (Source: GameBiz).  

By releasing official, commercial voice models of his speaking and singing voices, Kaji is providing a legal way for creators to use his talents. He previously explained that he started the project because he found himself “worrying every day” about the widespread usage of generative AI creating a legal gray area when it comes to rights relating to your voice (source: Robotstart). Generative AI, especially unauthorized use of human voices to train AI models, has become a contentious topic among Japanese voice actors. Last month, a group of 26 top Japanese anime and game voice actors formed the “No More Unauthorized Generative AI” group, with the aim of spreading awareness about this issue.