Split royalties on AI-generated music using my voice

• Electronic music producer and singer Grimes takes a stance on generative AI in music.
• “If an AI-generated song using my voice is successful, we’ll split 50% of the royalties,” she tweeted.
• The genre-crossing artist and former Elon Musk girlfriend said the technology could open-source art.

While the music industry considers the potential disruption wrought by generative AI, Claire Boucher (also known by her stage name Grimes) is embracing whatever the technology has to offer. And it seems.

The Canadian producer and singer, who earned critical acclaim more than a decade ago with her genre-defying album Visions, said Sunday night: twitter She has pledged to happily split 50% of the royalties on AI-generated songs that make good use of her voice.

This 50/50 split seems to be the same deal she ties in with other artist collaborations, robot or not.

But her statement didn’t end there. Boucher basically encourages the use of his voice in AI-based music projects, stating, “Use my voice freely without penalty. I have no label and no legal bindings.” I’m here.

She didn’t elaborate on exactly how such a royalty split would work, and it’s possible that someone may have the music from Grimes’ past albums on a record label that still holds the rights to the work. “Visions”, for example, was released on 4AD, a record label owned by the Beggars Group. A business associate for Grimes did not immediately return a request for comment. 4AD also did not immediately comment.

But while her proclamation runs counter to what major labels want, which is to receive some use of copyrighted music, Boucher has experimented with new technologies not only in her music but also in her personal life. has a long history.

Last month, she claimed that a brain-computer interface allowed her to use her brain to control a computer mouse.

The emergence of AI-generated music that mimics the vocal melodies and production styles of popular artists is treading legally questionable waters. Taking part of the original recording of an existing song and using it to train an AI model to create a new one is an example of copyright infringement.

Even before an AI-generated song featuring the voices of Drake and The Weeknd went viral last week, major tech companies such as Google were working on their own tools that could create music based on textual descriptions. .

“I think it’s cool to merge with machines, and I like the idea of ​​open sourcing all art and eliminating copyright,” said Elon Musk, the creator of Tesla and the father of two children. Boucher, a co-parent, insisted. “I’m just curious about what happens and I’m interested in being a guinea pig.”