Japan Actors Union Sets Generative AI Rules for Vocal Audio
The union's rules put written consent, notification, and compensation on the table for a field where voice actors have already heard their own voices reused without their say, and it treats animation and dubbing as special cases.
Reporting from 1 source: Anime News Network.
The Japan Actors Union presented a 21-page guidelines document on Thursday covering generative AI use in vocal audio work. It runs six sections and 22 articles, from prior notification and written consent to compensation, data oversight, and cease and desist requests. At the press conference, voice actors Masako Nozawa and Bin Shimada described hearing their voices used without permission.
The document defines vocal audio broadly, covering every recording of an actor or voice actor, private recordings included. It also defines AI use broadly: voice synthesis, deep learning, machine learning, prompt-driven modeling, metadata configuration, GUI operation, and file specifications all fall inside it.
Section 4 carries special provisions for Japanese-language vocal audio in animation and dubbed content, and for multilingual localization and overseas distribution. Others cover consent periods, prohibitions on penalizing actors who refuse or withdraw consent, and rights to request crediting and removal. Masako Nozawa, who voices Son Goku, said voices are being used without consent. Bin Shimada said hearing a generative AI version of his Broly voice singing Ado's "Usseewa" left him stunned, then angry.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.
Sources
- Anime News Network Interest Japan Actors Union Develops Guidelines for Generative AI Use