Google is negotiating with Universal Music Group, one of the world’s largest owners of record labels and artists’ contracts, to use artists’ images and voices in recording and videos of songs written by AI. 

The rise of generative AI has led to an outbreak of “deepfake” videos and recordings of songs created by AI but that borrow the voices and likenesses of human artists to perform them, often without their knowledge or consent or any form of payment.

Frank Sinatra’s voice was borrowed for an updated rendition of “Gangsta’s Paradise” and Johnny Cash’s baritone now sings “Barbie Girl.”

An AI recreated the voice of rapper Drake for a song earlier this year; after the recording went viral, Drake said it was “the last straw.” Rapper Ice Cube has called voice theft “demonic.”

The goal of Google’s discussions is to create an AI product that will enable fans to use artists’ voices and images to enact songs that a fan might choose—an original composition or a favorite by another artist—and to compensate the copyright holders. Artists would need to opt in, people familiar 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,” Jeffrey Harleston, Universal’s general counsel, said in Congressional testimony in July.

After years of battling with YouTube, the music industry now collects about $2 billion a year in royalties from videos using copyrighted tunes. The agreement lays a precedent for current discussions around AI.

Google also is in discussions with Warner Music, the third largest recording label, about a product, the Financial Times reported.

TRENDPOST: An artist usually protects not only her or his voice or likeness, but also their body of work, which also represents them in the world. Few well-known artists are likely to license their identities for random uses by pranksters and wannabe’s. Google’s product will find most demand among advertisers wanting to borrow a famous face or voice.