Japan court recognizes AI-cloned voices can fall under publicity-right protection

The Tokyo District Court has indicated that a person’s voice can fall within the scope of Japan’s publicity rights where it has customer-attracting value and is used primarily to exploit that value, in a significant ruling involving AI-generated imitation of Japanese voice actor Kenjiro Tsuda.

The court nevertheless dismissed Tsuda’s claim against TikTok’s operator because the account and videos at issue had already been deleted, meaning there was no longer an obligation to remove the material.

The decision is significant because it is being reported as the first Japanese judicial decision to expressly recognize that a person’s voice can fall within the scope of publicity-right protection. Tsuda’s attorney, Takeru Hirano, described the ruling as the first to explicitly place a voice within the scope of the right. Kyodo News similarly reported that the decision represented Japan’s first judicial determination concerning protection of voice rights.

AI-generated imitation of Tsuda’s voice

Tsuda, known for roles including Kento Nanami in Jujutsu Kaisen and Seto Kaiba in Yu-Gi-Oh!, sued the operator of TikTok after an anonymous account posted more than 180 videos featuring narration that his lawyers alleged had been generated using AI to imitate his distinctive voice.

The account reportedly began posting the videos in July 2024. Its content included videos about urban legends and other subjects, accompanied by narration that Tsuda’s lawyers said closely resembled his voice. The account used an image resembling a character from Jujutsu Kaisen voiced by Tsuda.

Tsuda’s lawyers argued that the unauthorized imitation infringed his publicity rights and sought removal of the videos from TikTok.

TikTok disputed the characterization of the voice, arguing that the narration was a “generic” or “universal” male voice and that any perceived similarity to Tsuda’s voice was subjective. The platform also argued that the voice itself was not responsible for attracting viewers to the videos.

Court addresses publicity rights

Although the court dismissed Tsuda’s claim because the relevant material had already been removed, it addressed the underlying publicity-right issue.

The court indicated that a voice can have customer-attracting value and that unauthorized use can constitute an infringement of publicity rights where the voice is used primarily to exploit that value. The judgment also reportedly characterized a person’s voice as a symbol of individual personality in the same way as a portrait.

Japan’s publicity rights are not established in a standalone statute. Instead, their protection has developed through case law concerning the commercial value associated with a person’s identity.

The Tokyo ruling therefore does not create a new statutory property right in every person’s voice. Rather, it provides an important judicial indication that an individual’s voice can, in appropriate circumstances, constitute an aspect of identity possessing the customer-attracting value protected by publicity rights.

Account had already been deleted

The court’s dismissal turned on the status of the disputed material.

According to Tsuda’s lawyer, the account operator deleted the account and its videos during the proceedings, around May 2026. As the material was no longer available, the court found that there was no remaining obligation requiring TikTok to remove it. Tsuda’s lawyers said after the judgment that the deletion had achieved the original objective of the lawsuit.

The case therefore did not result in an order requiring TikTok to remove the AI-generated videos. Its significance instead lies primarily in the court’s treatment of the voice itself under Japan’s publicity-right framework.

Japan had already addressed AI-generated voices

The judgment follows a significant development at the Japanese Ministry of Justice. In August, the Ministry published a report examining civil liability for unauthorized use of portraits, voices, and other aspects of a person’s identity, including issues arising from generative AI. The study examined the application of existing publicity-right and personality-right principles, as well as potential claims under Japan’s Unfair Competition Prevention Act.

The Ministry’s report did not establish a new standalone statutory right protecting a person’s voice. Instead, it provided an interpretation of how existing law and case law may apply to unauthorized uses of voices and likenesses generated or reproduced using AI. The Tokyo District Court decision now provides a judicial development alongside that government analysis.

Implications for AI voice cloning

The case highlights a growing legal issue for performers, brands, platforms, and AI developers: whether existing personality and publicity-right frameworks can protect commercially valuable aspects of an individual’s identity when they are reproduced by generative AI.

The issue is particularly significant for voice actors and other performers whose professional identities may be closely associated with distinctive voices.

AI voice-cloning technology can reproduce characteristics of a performer’s voice without reproducing an existing recording of that performer. The legal question is therefore not necessarily whether a particular recording has been copied, but whether the unauthorized imitation exploits the commercial value associated with the individual’s identity.

The Tokyo ruling does not establish that every AI-generated voice resembling another person will infringe publicity rights. Rather, it identifies circumstances in which a voice may fall within the existing publicity-right framework — particularly where it possesses customer-attracting value and is used primarily to exploit that value.

The decision also leaves questions around how similarity between a real person’s voice and an AI-generated imitation should be established, an issue that featured prominently in the litigation. Tsuda’s legal team reportedly used analysis based on multiple deep-learning models to compare the voices and submitted the relevant program code as evidence.

For performers, rights holders and online platforms, the ruling adds an important judicial development to Japan’s emerging approach to AI-generated identities — and suggests that the commercial value of a distinctive human voice can, in appropriate circumstances, receive protection under publicity-right principles.

← Back to News