Recently, audio engineer Young Guru went viral for voicing his concerns with artificial intelligence voice filters. He was reacting to a specific social media video of a man using a Kendrick Lamar AI voice filter. The software, Tacotron 2, allowed the man to do a spot-on replication of Kendrick’s voice. Pretty neat, huh? Not according to Young Guru and many others who are concerned with the ethical and legal ramifications of said tech.
There are currently no legal protections for public figures from folks using replication software to mimic their voices. Using it in the confines of your own home studio is one thing, but to create and market a work featuring a synthesis of someone’s actual voice is another. Producer Teddy Riley was accused of that after the release of Michael Jackson’s first posthumous album, Michael, back in 2010. Upon the album’s release, some fans cried foul when hearing the vocals on some of the project’s songs, claiming it was an MJ impersonator, and not the King of Pop himself. Riley placed the blame on the vocal processing software he used, Melodyne, while vouching for the authenticity of Michael’s voice on the album. While this is not an example of AI, it does show the potential duplicity that may arise from the use of speech altering software.
Midler vs. Ford Motor Company
In 1986, Ford Motor reached out to Bette Midler to sing one of her songs, “Do You Want To Dance,” for a campaign advertising the Mercury Sable. When Midler declined, the company hired a singer to impersonate her singing style, potentially leading consumers to believe they were actually hearing Bette. She took legal action against Ford for using her distinctive voice without her authorization. The appellate court ruled in her favor, stating “A voice, or other distinctive uncopyrightable features, is deemed as part of someone’s identity who is famous for that feature and is thus controllable against unauthorized use.”
The Future…
Advancements in technology have revolutionized the music industry several times over. We have the ability to replicate practically every instrument using commercially available hardware and software. It was only a matter of time before technology caught up and allowed creators to produce convincing approximations of not only the human voice, but specific voices of actual people. The aforementioned Midler case, while unprecedented, may hold little weight in regard to AI generated voices. Should we, as Young Guru believes, be proactive and lobby for lawmakers to implement legal protections for artists who may fall victim to this AI voice tech? Or, do we wait until a “new” 2Pac album hits the market to assess how to deal with this from a legal standpoint? Speaking of, how ’bout a Biggie/2Pac collab project? I can see it now: The 2Pac and Biggie Duets Vol. 1: East/West Summit. Hey, it wouldn’t be the first time the music industry employed questionable tactics to milk the legacy of dead artists.
Onward and upward…