With generative AI becoming more and more of an issue for creatives in all industries, voice actors in Japan are finding their own way to fight back against the unauthorized use of their voices. Last week, the union representing Japanese voice actors issued its own guidelines regarding the use of AI, as a number of actors spoke about the personal and professional challenges they’ve faced due to the technology.
Voice actors in Japan have been taking a stand against the unauthorized use of their voice for a number of years. In 2024, a group of voice actors came together to create their own organization protesting AI voice misuse, specifically citing the lack of legal copyright protections. One voice actor, Kenjiro Tsuda, known for voice roles in Attack on Titan, Chainsaw Man, and Yu-Gi-Oh!, even sued TikTok regarding the proliferation of videos using his voice on social media, though the case was dismissed once the videos were removed. While some voice actors have advocated for systems allowing for private, licensed use of AI to replicate performers’ voices, the general view of the technology among voice actors in Japan is that it could steal their work and future opportunities.
At a press conference on Thursday, the Japan Actors Union (JAU) presented the first version of its new guidelines regarding the use of generative AI. The guidelines define these new rules as covering any audio recordings made with AI technology replicating actors’ voices, including unauthorized use for voice synthesis, deep learning, machine learning, and so forth.
As reported by Automaton, the guidelines call for anyone seeking to replicate the voice of an actor to obtain consent at every step of the process. Should an AI company wish to use an actor’s voice in AI training, it must obtain permission from the performer. Also, the authorized use of AI-generated voice is categorized as a type of “performance,” and companies will be required to pay salaries and/or royalties to actors.
At the conference were a number of voice actors who spoke against the use of the technology for unauthorized purposes. Masako Nozawa, best known for voicing Goku in the Dragon Ball franchise, told Japanese news broadcaster NHK, “I can’t forgive them for using it without my permission. It makes me really angry, it’s no different than theft. My voice is mine, as are my performances, and if that’s going to be used, even a little bit, they should ask for permission.” The voice actor also claimed that an AI version of her voice that she’d heard “lacked heart.”
The union represents roughly 2600 voice actors and actors, and the guidelines are among the first to be released in Japan regarding the use of AI for voice recordings. It sends a strong signal against AI in a country where its use is generally more accepted than it is in the U.S. And given the relative strength of Japan’s unions compared to those seen in the U.S., it’s likely this guidance will have an impact on the unauthorized use of the technology to replicate actors’ voices. Japan’s government is currently facing questions on how to deal with generative AI in response to both cyberattacks and copyright infringement, making this announcement from the JAU a notable one as the country grapples with how to address AI technology.