To America, where a group of high-profile voice actors, many with a firm footprint in the anime world, have declared war on artificial intelligences. Or rather, artificial intelligence’s meat masters, noting that many actors performing “voice performance replication work are unaware of or do not fully understand their rights regarding employment contracts.”

Vocal Variants, a pressure group including Yuri Lowenthal, Stephanie Sheh and Matt Waterson, outlined a set of simple demands and stipulations. The group aims to inform companies and actors, chiefly to warn actors off inadvertently signing away all their rights to exclusivity in their own voices.

Any voice actor with a significant body of work inadvertently creates an audio bank of their voice. This is particularly true in the gaming world, where actors are often less reciting a script than delivering “barks” and soundbites. Sakura Wars even made a big deal out of its audio component, for which the Japanese voice actresses recorded themselves saying all 100 syllables in the language, thereby allowing the audio software to address the (Japanese) player by name, and, in theory, to say anything the computer wanted.

Give an AI enough material to play with, and it can generate dialogue as if the actor is saying it themselves. It was Steve Blum, the voice of Cowboy Bebop’s Spike Spiegel, who first poked his head above the Twitter parapet to ask who had the right to make him say things he never said. It started up a social media storm that led to the formation of Vocal Variants, and its statement that voice actors were entitled to safe storage of their voices, clear stipulation of what those voices might be used for, approval on the use of their voices to generate synthetic dialogue, and appropriate payment for use.

Stephanie Sheh noted on Twitter that although many of the AI apps agree to take down unsanctioned audio files, the actors are often obliged to police them themselves, and often don’t even know their voice has been uploaded unless they join each specific app service.

“As AI/Synthetic voice work now covers much uncharted territory,” says the Vocal Variants website, “it’s imperative that we collaborate to create and amend laws and contracts to protect both laymen and professional performers against deep fakes, improper use and exploitation of recorded performances.”

Jonathan Clements is the author of Anime: A History. This article originally appeared in NEO #229, 2023. Since its publication, Vocal Variants has become a division of the National Association of Voice Actors.