In a major push to safeguard performers’ livelihoods, the Japan Actors Union has drafted strict new guidelines demanding explicit, step-by-step consent and separate commercial compensation whenever companies or individuals attempt to replicate the voices of actors and voice actors using artificial intelligence.
The Bottom Line
- The Japan Actors Union, representing roughly 2,600 actors and voice actors, has established new operational guidelines requiring explicit consent before any human voice data is fed into an AI system.
- The rules mandate separate authorizations and distinct fee structures for every language translation and every individual usage phase, addressing unauthorized deepfakes and AI voice cloning.
- The push arrives alongside a pivotal Tokyo District Court ruling recognizing that unauthorized voice cloning can infringe upon a performer’s legal publicity rights.
Securing the Human Voice Against Unauthorized AI Training
The Japan Actors Union represents approximately 2,600 stage, screen, and voice actors who are increasingly confronting unauthorized artificial intelligence clones circulating online. These synthetic voice models frequently match the exact vocal cadence of well-known performers without their knowledge or financial compensation. Under the newly drafted guidelines, users must secure express permission before taking any action with a performer’s vocal data. This requirement applies right from the initial step of feeding human voice recordings into an AI system to build an artificial voice model.
The Japan Actors Union (JAU) has officially announced its support for voice actor Kenjiro Tsuda's lawsuit against TikTok over AI-generated videos allegedly imitating his voice without permission"The Japan Actors Union supports Kenjiro Tsuda's courageous actions." pic.twitter.com/Ke8i3Io2ZE
Furthermore, separate approvals are required for generating subsequent speech files, embedding them into videos, and publishing the final media. Companies must explicitly state the intended commercial application for every single deployment. Yuko Sasaki, a board member of the union and renowned voice actor known for her role as the grandmother in the anime series Chibi Maruko-chan, emphasized that performers generally harbor no hostility toward artificial intelligence technology itself. Instead, the primary objective is preventing unapproved usage that bypasses the talent entirely.
Back in February 2026, the union issued early warnings urging actors, talent agencies, and production companies to reject contracts featuring ambiguous clauses regarding AI utilization. Those warnings stressed that once voice data is ingested into an AI training dataset, it is technically difficult to completely remove. To address this operationally, the union partnered in November 2025 with the trading firm Itochu and its IT subsidiary Itochu Techno-Solutions to build “J-VOX-PRO,” a specialized voice database. The platform allows performers to register their voice profiles, establish specific usage conditions, embed digital watermarks, and secure legitimate paid contracts.
Navigating Multi-Language Localization and Deceased Performers
Advanced voice-cloning technology now enables companies to translate original vocal performances into foreign languages while entirely preserving the distinctive acoustic characteristics of the original speaker. To close this loophole, the union’s new framework mandates separate consent and independent remuneration for every individual language adaptation. Granting permission for a voice model in one specific market or language does not grant automatic rights for regional expansions.
The union has also established precise protocols regarding the digital recreation of deceased actors and voice actors. Talent agencies representing performers who have passed away have raised significant concerns regarding unauthorized commercial exploitation. Under the new guidelines, any entity seeking to synthesize a deceased performer’s voice must first obtain explicit approval from surviving family members or authorized estate representatives while honoring the documented wishes of the artist.
| Milestone / Initiative | Timeline | Key Stakeholders | Core Objective |
|---|---|---|---|
| J-VOX-PRO Database Announcement | November 2025 | Japan Actors Union, Itochu, Itochu Techno-Solutions | Provide a secure platform for authorized voice registration and digital watermarking. |
| Contract Warning Issued | February 2026 | Japan Actors Union, Agencies, Production Houses | Warn industry members against signing ambiguous contracts permitting unrestricted AI training. |
| Tokyo District Court Ruling | September 30, 2026 | Kenjiro Tsuda vs. TikTok Operator | legally recognized that unauthorized voice cloning can violate personality and publicity rights. |
| Ministry Working Group Proposals | April – July 2026 | Japanese Ministry of Justice | Investigate legal protections surrounding unapproved voice and image replication. |
Legal Precedents and the Limit of Voluntary Industry Standards
The urgency behind these guidelines was underscored by a notable legal development on September 30, 2026. Prominent Japanese voice actor Kenjiro Tsuda filed a lawsuit against the operator of TikTok after an anonymous user account uploaded more than 180 videos utilizing an AI-generated voice nearly indistinguishable from his own. The Tokyo District Court ruled that unauthorized voice simulation can constitute a legal infringement under publicity rights, affirming that a person’s voice forms a core component of their personal identity.
While Yuko Sasaki welcomed the ruling as a vital step forward for legal recognition, she cautioned that court decisions alone will not resolve the widespread proliferation of internet deepfakes. Simultaneously, the Japanese Ministry of Justice convened a specialized working group between April and July 2026 to evaluate legislative options regarding unauthorized voice and image exploitation. Although that group reviewed existing frameworks—including laws against unfair competition and the protection of personal identity—it stopped short of drafting a dedicated standalone statute protecting human voices.
Consequently, the Japan Actors Union is pressing studios and tech platforms to adopt its newly minted guidelines as standard business practice, while continuing to lobby lawmakers for enforceable statutory protections and stricter copyright rules governing AI training data.
