Photo/Illutration Voice actor Mika Kanai at a news conference on Nov. 19 (Ryo Jozuka)

Amid rapid advances in artificial intelligence, the private sector is moving quickly to protect the voices of actors and voice actors while promoting their commercial use.

At a news conference in Tokyo on Nov. 19, voice actor Mika Kanai recorded about 20 seconds of dialogue on a smartphone to demonstrate new AI-powered voice conversion technology. 

About 20 minutes later, the audio was played for the audience at the venue after being converted into languages such as English, Chinese and Hindi.

The husky timbre and other distinctive qualities of Kanai’s voice remained intact.

The technology can convert voices into more than 30 languages, according to the voice protection and multilingualization association, which organized the event.

The association, established by a voice actor agency and other parties, said it is also technically possible to convert voices of overseas actors into Japanese.

The Japanese subsidiary of ElevenLabs, a U.S. AI voice synthesis company known for detecting fake videos during the U.S. presidential election last year, provided technical support.

The technology can guarantee authenticity by embedding sounds called “digital watermarks” in frequencies inaudible to the human ear.

“If we can harness the power of anime voice actors for international expansion, the reach of Japanese anime will grow even further,” said Masakazu Kubo, the association’s representative director.

Last year, the market of Japan’s anime industry reached 3.8 trillion yen ($24 billion), double what it was a decade earlier, according to a survey by the Association of Japanese Animations.

Overseas sales accounted for more than half that figure, increasing 6.6 times from 10 years earlier.

If AI can render voice actors’ performances into multiple languages, exporting works will become easier.

New business opportunities are also expected, such as anime merchandise with voice features and narration services in various languages.

However, Japan currently lacks a legal framework to directly protect “voice rights.”

As unauthorized use of voice actors’ voices mimicked by AI becomes a growing issue, private companies are moving to license their voices and turn them into legitimate products.

The Japan Actors Union announced on Nov. 14 that it will build a commercial database with trading house Itochu Corp. by training AI on voiceprints and speech characteristics of participating actors and voice actors.

The project will detect unauthorized use of registered voices and match participating performers with companies and organizations that want to use their voices for AI and other purposes.

In October, NTT West Inc. announced plans to launch a business that issues certificates for AI-generated audio using voices of actors and voice actors, thereby generating licensing fees.

The regional telecommunications carrier is targeting sales of 1 billion yen by 2028, 10 billion yen by 2030 and 100 billion yen within 10 years.

While expectations are growing for the commercial opportunities that AI-powered voice conversion technology could bring, particularly in the anime industry, many voice actors have raised concerns that it could replace them.

That is because the traditionally vital work of dubbing foreign films and anime could become unnecessary.

“We could lose our ‘dubbing culture’ because overseas productions can be turned into Japanese with AI,” said Ayano Fukumiya, a voice actor, who serves as secretary-general of the Nippon Anime and Film Culture Association.

Related News