Summary
The Japan Actors Union, a cooperative representing about 2,600 actors and voice performers, has issued voluntary guidelines for generative voice AI. The union says producers and AI developers should explain the purpose and obtain a performer’s advance permission at three stages: training, generation, and publication or use. Performers should be able to set the scope of use, while contracts should state the conditions clearly. The guidelines classify use of a generated voice as equivalent to a new performance and call for compensation based on actual use and reach. AI dubbing into additional languages would require new permission and extra payment tied to international distribution. The union plans to promote the standard and seek statutory protection for voice rights in Japan.
Hana’s perspective
I think the guideline’s strongest feature is that consent is not treated as a one-time permission to train a model. It separates training, generation and release, then connects each stage to scope and payment. That structure fits animation and dubbing workflows, where one recorded performance can otherwise be reused across characters, languages and territories far beyond the original job. Calling synthetic use a new performance also gives producers a clearer budgeting principle. The guideline is voluntary, so adoption and enforcement remain open questions, but it offers a concrete contract model while Japan debates broader legal protection for voice.
Are you a writer with a great story? We'd love to hear from you: APPLY HERE