METI Drafts Rules to Stop Unauthorised Commercial Use of People’s Faces and Voices as Generative AI Spreads
Japan’s Ministry of Economy, Trade and Industry (METI) presented an expert panel on November 13 with a concrete draft of the acts that a new category of unfair competition covering the unauthorised use of a person’s likeness or voice would target, the Yomiuri Shimbun reported. The proposal, made at the 32nd meeting of the unfair competition prevention subcommittee, responds to generative AI making it easy to create convincing copies of someone’s face or voice, and to the spread of such copies in products, advertising and videos without permission.
The draft targets three types of commercial use that trade on the appeal of a person’s likeness or voice: selling the likeness or voice itself as something to be enjoyed, such as photo cards or voice messages; using it to differentiate a product, such as T-shirts printed with a face or the narration of an audiobook; and using it in advertising. According to YUKU NEWS, the three types follow the Supreme Court’s reasoning in the Pink Lady case. Impressions and voice mimicry would be excluded, as would people who simply happen to look or sound like someone else using their own face or voice, so that legitimate expression is not caught. An impression could still count as an infringement if it is used in a way that makes people think it is the real person.
What it means in practice
Under the draft, the right to sue would basically sit with the person concerned; a talent agency could be allowed to bring a case only if it holds an exclusive licence from the performer. At the meeting, one attendee argued that using a likeness or voice to make money on social media should also be covered. Panel member Izumi Hayashi, a lawyer, welcomed the fact that protection would not depend on how famous someone is, but said impersonation and sexual deepfakes cannot be fully stopped by the Unfair Competition Prevention Act, and called for a special law from the ministries concerned and a procedure to ask platforms to remove content. METI has been considering a revision because existing law is seen as unable to deal properly with unauthorised use of likenesses and voices, and it will continue the panel debate before designing a detailed system.
For advertisers, game makers and voice-AI companies, the practical message is to review rights clearances and in-house rules for generative AI now. The proposal builds on steps already taken by performers, such as the Japan actors union’s consent and pay rules for AI use of voices, and on wider rule-making, including the government’s review of the AI Act, which is weighing penalties. What to watch is whether social media monetisation is added to the scope and when a bill reaches the Diet.