On October 2, IT Home relayed AFP: on September 30, local time, a Tokyo court accepted voice actor Kenjiro Tsuda’s claim in his case against TikTok. The court found that an AI imitation of his “thick” baritone without permission involved an infringement of rights, and for the first time in Japanese judicial practice confirmed that a person’s voice is legally protected. The judgment says that using an actor’s voice without the actor’s permission should be treated as an infringement of publicity rights. The report says this is also the first time Japan has heard a case of this kind.

[1]
A blank film strip: the left frame is circled in red, and the right frame is torn open to white.
Every frame is empty. A red circle marks the left frame, a whole empty frame sits in the middle, and the right frame is torn open. That is a voice brought inside protection, with the takedown left empty. An illustration, not a court record., AI-generated illustration, not a news photograph

Tsuda has worked as a voice actor for more than 30 years, on a large number of films, animated series, and video games. The report says one of his best-known roles is Seto Kaiba in Yu-Gi-Oh. He said an anonymous TikTok account, from July 2024 through the second half of last year, uploaded videos that used an AI clone of his voice as narration. He said the narration had his distinctive low, “thick” voice, enough to judge that it was an imitation of him. His lawyers said the account once had more than 200,000 followers, and that it drew attention because a voice like a well-known actor’s was narrating “suspicious, vulgar content.”

[1]

The court accepted the claim only in part. The account had already been deleted, so the court did not grant his request that TikTok take the videos down. Confirming that a voice is protected is not the same as those videos still being removed.

The court record, as relayed, says TikTok argued that the videos used only a “general male voice,” and that whether it really resembled Tsuda was disputed. The report does not describe the comparison method the court used, and it does not state damages.

[1]

The report adds background that is not the holding in this judgment. Japan’s government promotes AI as national strategy, while anime voice actors increasingly worry that their work and personal likeness are used to train models and that protection is still thin. The article says that last year the Japanese government asked OpenAI not to use anime and manga, described as “irreplaceable treasures,” to train the Sora 2 model in a way that infringes copyright. That is a separate demand, not the text of Tsuda’s judgment.

[1]

要点

  • On September 30 a Tokyo court found that an unauthorized AI imitation of Kenjiro Tsuda’s voice involved an infringement, and confirmed that a voice is legally protected.
  • The court treats use of an actor’s voice without permission as an infringement of publicity rights. The report calls this Japan’s first case of the kind.
  • The account was already deleted, so the court did not order the videos taken down. TikTok said it was only a general male voice.
  • The request to OpenAI about anime and manga training is background in the report, not the holding in this judgment.