Google and Universal Music Want to License Voices Used in AI Music

Google and Universal Music Want to License Voices Used in AI Music

As generative AI balloons, which have allowed anyone to create deep fake music, Google and Universal Music are reportedly working on modalities to license the voices and melodies of musicians used for AI-generated music. 

The arts industry, especially music has been struggling to develop a workaround when it comes to AI-generated music and some artists have openly expressed their displeasure at the use of AI to mimic their voices. This comes as there has been a rise in people using generative AI to create music, images and videos using established artists’ voices, styles or sounds without their permission.

Rapper Ice Cube openly blasted the practice calling it “demonic” and that perpetrators must be sued. This is not the first time that Universal Music Group (UMG) is stepping in to protect music copyrights this year as in April the company wrote to streaming services including Spotify and Apple Music, asking them to prevent AI companies from accessing their libraries.

A relief to artists?

Talks between the two companies could be a relief to artists, as they may get something whenever their voices and likeness are used to generate music by AI.

Although still in the early stages, the discussions between the two companies have been confirmed by four sources close to the developments, according to a report by Financial Times.

While there is no specific product yet that is expected to be launched soon, the ultimate goal is to come up with a platform or tool that allows fans to create the songs but pays owners of the copyrights for it.

This will also address what has become one of the problems with generative AI that is deep fakes going viral.

One of the most famous deep fake AI-generated songs is “Heart on my sleeve,” a song that used AI-generated vocals that imitated Drake and The Weeknd. The song went viral on social media platforms such as TikTok. The track was later pulled from streaming platforms for copyright infringement.

Another example is “Gangsta’s Paradise” which features Frank Sinatra’s voice while Johnny Cash’s voice was used in a pop single “Barbie Girl.”

According to the Financial Times, a user account by the name PlugginAI on YouTube offers tracks that mimic the voices of the late Tupac and Notorious B.I.G.

Also read: Ripple’s CLO Urges Probe into SEC’s Bill Hinman Amid Conflict Suspicions

Music executives agree

Universal Music general counsel Jeffrey Harleston told US lawmakers his displeasure at the practice of capitalizing on artists’ voices.

He said: “An artist’s voice is often the most valuable part of their livelihood and public persona, and to steal it, no matter the means is wrong.”

The third largest music label Warner Music has also been in talks with Google for a product, according to sources close to the developments as per the Financial Times report.

Warner Music CEO Robert Kyncl told investors on Tuesday, August 8 that with the right framework in place, AI could “enable fans to pay their heroes the ultimate compliment through a new level of user-driven content…including new cover versions and mash-ups.”

The increase in free-to-access tools, which generate videos and audio content, for instance, Midjourney means the average user can create deep fake music, a situation that has left artists unhappy.

According to the Telegraph, the obtaining situation brings back memories of the rise of Google’s YouTube, where copyright infringement by its users resulted in a protracted dispute with the music industry, although the site now pays $2 billion a year in royalties to record labels.

Artists have a choice

Kyncl said artists should have a choice when it comes to AI music adding his company’s priority would be to ensure artists have a choice to opt in.

“There are some that may not like it, and that’s totally fine,” he said.

Electronic artist Grimes has offered to allow people to use her voice but split the royalties.

Gunnercooke LLP intellectual property partner Rosie Burbidge said: “If the music itself is AI generated then we are in clearer copyright infringement territory and, provided it is possible to prove that the AI generated music was trained using copyright works and there are similarities in the music or lyrics, there is a greater ability to prevent such use as it is likely to be a copyright infringement.”

Recently, musician Ed Sheeran added his voice to the debate on AI in music saying he is “not buying into the AI hype,” adding his fears of massive job losses due to the technology.

“The whole point of society is we do jobs and do things, if everything is done by robots, everyone is going to be out of work,” he said.

He added he finds AI “a bit weird,” and that society does not need it.

“I’m just like ‘have you seen movies where robots kill us?’”

Image credits: Shutterstock, CC images, Midjourney, Unsplash.