Japan’s actors’ union has proposed guidelines that would require consent for generative AI use of performers’ voices, with compensation for AI-generated replicas. The proposals also set conditions for translated performances and the use of deceased actors’ voices, giving the debate a practical shape beyond the familiar demand to “protect artists”.
Watch Desk analysis
What happened
The Japan Actors Union’s guidelines cover four areas, according to Eurogamer, citing Automaton. They call for permission at every stage of generative AI use, including model training; treat AI replicas as new performances eligible for royalties or compensation; and require permission for AI-powered translation, which should preserve the original performance’s artistic integrity.
For deceased actors, the guidelines propose a ban on AI use unless the performer gave express permission during their lifetime. Even then, they say permission from the actor’s family and affiliated agency or company, as well as royalties, would be required. The union says it wants to encourage industry-wide adoption and government legislation. These are proposed guidelines, not binding law.
Why it matters
Voice models can reproduce a performer’s distinctive work, while translation systems can carry it into new languages and markets. The union’s proposals treat those uses as decisions requiring permission and, in some cases, payment, rather than as a technical afterthought.
The distinction between guidance and law matters: the proposals do not, by themselves, compel AI companies to follow them. But they offer a concrete position for an industry negotiating over consent, compensation and who gets to authorise a voice’s reuse.
Our read
This is a substantive attempt to turn “respect performers” into rules people can argue over. The most consequential test will be whether studios, AI companies and lawmakers adopt the principles, especially the requirement for consent at each stage rather than a single broad permission slip.
What to watch
- Whether the union publishes further detail on how consent and royalties would work in practice.
- Whether entertainment companies or AI developers adopt the proposals.
- Whether Japanese lawmakers take up the union’s call for legislation.
Discussion spark: Should AI voice use require performers’ consent and payment for every new use, or can a single broad licence fairly cover training, replicas and translation?
Sources and evidence
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.