Particle.news

Japan Panel Says Voices Are Legally Protected Against AI Deepfakes

The advisory panel's draft signals that commercial AI voice likenesses can be blocked or yield damages with a final report due in early August.

Overview

  • The Ministry of Justice advisory panel provisionally approved a draft report on Monday that lays out civil liability for unauthorized AI-generated uses of voices and images.
  • The draft explicitly declares voice a protected personal attribute and applies the Supreme Court’s 2012 framework of a “right not to be used without permission” plus a commercial publicity right to voice likenesses.
  • The panel says specific uses can be unlawful, for example when an AI reproduces an anime character’s voice closely and the audio is posted on social media to earn revenue.
  • If publicity-right infringement is found, victims could seek removal of content and monetary damages, but voice actors warn that low pay per role and high litigation costs will limit many people’s ability to sue.
  • The move responds to more realistic generative-AI deepfakes that place words or songs in performers’ mouths and is meant to deter some abuses, though enforcement, compensation and industry practices will determine how much protection people actually gain.