Tokyo District Court Recognizes Voice as Potentially Protected Under the Right of Publicity in AI Voice Imitation Case

On September 30, 2026, the Tokyo District Court in Japan addressed a legal dispute involving AI-generated voice imitation, recognizing that a person's voice, like their likeness, may be protected under the right of publicity.

(Background)

Japanese voice actor Kenjiro Tsuda (津田健次郎) filed a claim against TikTok's operator, seeking the removal of videos that allegedly used an AI-generated voice resembling his own.

Between July 2024 and September 2025, the TikTok account in question had published 188 videos featuring urban legends, ghost stories, and other mysterious topics. Tsuda argued that the videos infringed his right of publicity and violated Japan's Unfair Competition Prevention Act. He sought their removal on the grounds that the unauthorized use of a voice resembling his own could mislead viewers and exploit his professional reputation.

(Key Findings and Legal Issues)

The Tokyo District Court recognized that a person's voice, much like their image or likeness, can represent an individual's identity and may therefore qualify for protection under the right of publicity.

The court explained that when a performer's voice has commercial appeal—the ability to attract customers or promote products and services—unauthorized use of that voice for the purpose of exploiting its commercial value may constitute an infringement of publicity rights.

This recognition is particularly relevant to voice actors, whose distinctive voices are often central to their professional identity and commercial value.

Tsuda's representatives argued that the AI-generated narration closely resembled his voice, potentially leading viewers to believe that he had personally narrated the videos. They maintained that the videos attracted audiences by taking advantage of his recognizable voice, thereby infringing his right of publicity. They also argued that the use of the voice could violate Japan's Unfair Competition Prevention Act, which prohibits certain practices that create confusion with another person's goods or business.

TikTok's operator disputed these claims, arguing that the voice was not generated using Tsuda's actual voice recordings. Instead, it maintained that the audio was a generic male voice created by training AI on a third party's voice and did not sufficiently resemble Tsuda's voice to constitute an infringement.

The dispute raised important legal questions: How closely must an AI-generated voice resemble a real performer's voice to infringe their rights? And when does the use of a synthetic voice amount to exploiting someone's commercial appeal? However, the court did not resolve these questions in this case. Why the Court Dismissed the Removal Request

During the proceedings, the account owner deleted the TikTok account on May 29, 2026, and did not request its restoration within the platform's 30-day recovery period. As a result, the court accepted that the videos had also been removed from TikTok's servers by approximately June 28, 2026.

The court concluded that even if TikTok's operator might previously have been required to remove the videos, there was no longer any content to remove by the time oral arguments concluded. It therefore dismissed Tsuda's removal request.

Importantly, the dismissal did not mean that the court considered the AI-generated voice lawful. The court did not determine whether the voice actually resembled Tsuda's, whether it exploited his commercial appeal, or whether the videos infringed his publicity rights or violated the Unfair Competition Prevention Act.

Tsuda's legal representatives highlighted the significance of the ruling, noting that the court had expressly recognized the possibility of protecting a person's voice under the right of publicity.

Intellectual property experts have also pointed out that several important questions remain unanswered, particularly how courts should assess similarity between AI-generated voices and real individuals, and what constitutes unauthorized exploitation of a voice's commercial value. These issues will likely require further clarification through future cases.

As AI voice-generation technology becomes increasingly sophisticated, the distinction between synthetic speech and recognizable human voices is becoming more difficult to draw.

This case highlights an emerging challenge for intellectual property law when a voice can be digitally recreated how should the law protect the identity and commercial value behind it?