On September 30, 2026, the Tokyo District Court issued a landmark ruling recognizing that a person’s voice can be protected under Japan’s right of publicity—the first judicial decision in the country to extend this protection beyond a celebrity’s name and likeness.
This ruling arrives at a moment when generative AI is forcing jurisdictions worldwide to reckon with the legal boundaries of synthetic media. Although the court ultimately dismissed the claim on procedural grounds, the decision establishes an important doctrinal foundation and sends a clear signal to AI developers, platforms and content creators operating in Japan.
The Dispute: AI-Generated Voice Content
The case was brought by a well-known voice actor (the “Plaintiff”) against a major short-video social media platform operator (the “Platform Operator”) over AI-generated imitations of the Plaintiff’s voice. The Plaintiff filed suit against the Platform Operator seeking to remove videos that used generative AI to imitate the Plaintiff’s voice without authorization.
An anonymous account posted the videos to more than 210,000 followers. Between July 2024 and September 2025, the account published at least 188 videos narrating urban legends and paranormal stories. Viewers commented that the narration closely resembled the Plaintiff’s distinctive voice. The account was deleted in May 2026, and the videos were removed from the Platform Operator’s servers before the court’s decision.
The Platform Operator argued that the poster stated on an external site that the voice was generated by training an AI model on a friend’s voice, and contended there was no violation of law or infringement of rights. It also argued that by deleting the account and all associated data, the Plaintiff lacked a cognizable interest in pursuing the claim.
Court Finding: Voice Protected by Right of Publicity in Japan
The presiding judge’s ruling addressed the scope of Japan’s right of publicity in the context of voice for the first time. The court held, as a matter of general principle, that:
- A person’s voice, like their likeness, is a “symbol of their personality” and is therefore capable of protection under the right of publicity.
- The unauthorized use of a performer’s voice infringes the right of publicity where the use is found to be aimed solely at exploiting the voice’s customer-attracting power —the same “exclusively” test established by the Supreme Court in the Pink Lady decision (2012) for name and likeness.
The court declined to reach the merits of whether the specific videos infringed the Plaintiff’s publicity rights. Because the videos had already been deleted from the Platform Operator’s servers, the court concluded that the Platform Operator had no further duty to remove them and dismissed the claim.
The judgment did not address the Plaintiff’s additional argument under the Unfair Competition Prevention Act alleging that the poster had earned JPY 500,000–750,000 per month by creating confusion with the Plaintiff’s voice.
How This Fits With the Ministry of Justice’s Interpretive Guidance
The Tokyo District Court’s decision is consistent with the interpretive direction signaled by the Japanese Ministry of Justice (MOJ). In April 2026, the MOJ convened a study group comprising intellectual property scholars and practicing lawyers to examine the legal protection of voice in the age of generative AI.
The group’s report from August 2026 concluded that:
- Voices that can identify a specific individual—such as a celebrity’s or a well-known performer’s voice—are subject to legal protection.
- Individuals possess a “right not to have their voice used without permission,” recognizing both personality-right and publicity-right dimensions of voice.
The court’s framing that a voice is a “symbol of personality” deserving of the same publicity-right protection as a name or likeness aligns squarely with the MOJ’s interpretive position. Together, the judicial and administrative developments suggest a growing consensus in Japan that existing legal principles extend to voice, even in the absence of dedicated legislation.
Open Questions About Infringing Content
While the decision is a significant milestone, it leaves several important questions unresolved:
- Proving similarity. Demonstrating that an AI-generated voice is “the same as” a celebrity’s voice is considerably more difficult than proving likeness or name misappropriation. The court did not articulate an evidentiary standard, and the threshold for infringement remains unclear.
- Platform liability. Because the videos were deleted before judgment, the court did not opine on the scope of a platform’s obligations when infringing content remains live—including notice-and-takedown duties or potential secondary liability for hosting AI-generated voice content.
- Scope of the “exclusively” test. Under the Pink Lady framework, publicity-right infringement requires that the use be aimed “exclusively” at exploiting customer-attracting power. How this test applies to AI-generated voice content that serves both entertainment and commercial purposes (e.g. monetized narration accounts) remains to be clarified.
- Unfair competition claims. The court’s silence on the Unfair Competition Prevention Act argument leaves open whether voice cloning can also be challenged under Japan’s unfair competition framework, a potentially broader and more flexible cause of action.
- Future legislation. Japan currently has no statute that expressly protects voice rights. While the court and MOJ have extended existing doctrines, industry groups such as the Japan Actors Union continue to call for dedicated legislative protections, citing the burden of litigation and the scale of harm—a recent survey by an NPO found over 40,000 social media posts suspected of unauthorized use of celebrity voices and likenesses, with an estimated economic impact of JPY 2–4.5 billion.
Legal developments protecting voices against AI imitation are emerging in multiple jurisdictions around the world. This decision marks Japan’s own contribution to that global trend.
What This Landmark Ruling Means
The decision and the MOJ’s interpretive guidance carry practical implications across the AI and entertainment value chain.
For AI Developers and Platforms
- Audit training datasets for voice data belonging to identifiable individuals, particularly celebrities and performers. Ensure that proper consent or licensing is in place before using such data to train voice-synthesis models.
- Review and strengthen content moderation and takedown processes. The court’s decision to dismiss on the basis that the content had been removed underscores the value of prompt action—rapid takedown can be a meaningful defense.
- Implement technical safeguards, such as voice-identification or fingerprinting tools, to detect and flag AI-generated content that may replicate the voice of identifiable individuals.
For Entertainment and Media Companies
- Update contracts with performers and voice actors to expressly address AI use of their voice, including the scope of permitted uses, training rights, synthetic reproduction rights and compensation terms.
- Develop internal policies governing the use of AI-generated or AI-modified voice content in productions, advertising and promotional materials.
For Brands and Advertisers
- Exercise caution when using AI-generated voiceovers that resemble identifiable celebrities. Even if the AI output is not an exact replica, the risk of a publicity-right claim increases if the resemblance is commercially exploited.
- Obtain explicit licensing agreements from talent before using AI-synthesized versions of their voice in marketing or advertising campaigns.
For Talent and Agencies
- Negotiate AI-specific clauses in engagement contracts that define how voice data may (and may not) be used, retained and commercialized.
- Monitor platforms for unauthorized AI-generated reproductions and act promptly to request takedowns. The case illustrates that deletion before judgment may moot a claim.
- Consider registering voice samples or maintaining documentation that can aid in proving similarity in any future enforcement action.
The Orrick team monitors developments in AI voice-rights law in Japan and globally. If you have questions about how this decision may affect your business, or if you need guidance on AI-related intellectual property, content moderation or talent contracting, please reach out to your Orrick contact or the authors of this alert.