Tokyo Court Rules Unauthorized AI Voice Cloning Violates Publicity Rights
A Tokyo court ruled in favor of veteran voice actor Kenjiro Tsuda, establishing Japan's first judicial precedent confirming that human voice is protected under legal publicity rights. The court found that an anonymous TikTok account infringed on Tsuda's rights by using an unauthorized AI-cloned version of his distinctive baritone voice for video narrations. This landmark decision provides crucial legal backing for voice actors and creators in Japan's major entertainment and anime industries against unauthorized generative AI voice cloning. It aligns with a broader international legal movement to expand personality and publicity protections to biometric data like voice models. Although the court recognized the infringement, it did not compel TikTok to remove the videos because the offending account had already been deleted. TikTok argued the videos used a generic male voice, but the court agreed with Tsuda that his low, recognizable tone was clearly imitated.
## BACKGROUND
AI voice cloning technology leverages deep learning models trained on audio samples to synthesize a person's pitch, cadence, and vocal identity. Legal systems globally are determining how to regulate this technology, as traditional copyright laws often protect written works or sound recordings rather than the intrinsic voice identity of an individual.