Cyberpunk 2077 Phantom Liberty used AI to replace a voice actor who died

Cyberpunk 2077 publisher and developer CD Projekt used artificial intelligence to recreate the voice of a deceased actor.

As reported BloombergCD Projekt used the technology to add new vocal lines by famous Polish voice actor Miłogost Reczek.

Reczek provided the Polish-language voice for the character Viktor Vektor in Cyberpunk 2077, but has since died in 2021.

As such, CD Projekt was left trying to figure out how to use the character in the game’s Phantom Liberty expansion.

One option was to get a new voice actor to play Viktor in the expansion, and have that actor re-record Reczek’s lines for the base game to ensure the character sounds the same across the entire experience.

However, as Mikołaj Szwid, director of localization at CD Projekt, explained in a statement to Bloomberg, the company “did not like this approach” because Reczek “was one of the best Polish vocal talents” and his performance was considered “excellent.”

Instead, the company got a new voice actor to perform new lines, then used voice reproduction software called Respeecher to change the voice and make it sound like Reczek was performing the lines.

CD Projekt says it obtained permission from Reczek’s family before going ahead with the plan, and Szwed added that the actor’s children “have been very supportive.”

The use of artificial intelligence in the entertainment world is currently a major point of contention. Some members of the acting community worry that it will be used in a negative way, as studios try to cut costs by using it to replace actors, or to alter their images or performances without compensating them fairly.

Last month, members of SAG-AFTRA, the labor union for American actors, voted overwhelmingly in favor of a potential strike in the video game industry.

Negotiations have so far failed because of what SAG-AFTRA calls unacceptable terms on “some of the issues most important to our members, including wages that keep pace with inflation, protections around exploitative uses of artificial intelligence, and basic safety precautions.”