Google is in talks with major music labels to license artists’ voices and tunes for AI-generated songs, according to a report.
The tech giant has discussed the prospect with Universal Music and Warner Music, the Financial Times said.
It could reportedly see a tool developed allowing fans and creators to make tracks using recognisable vocals and melodies legitimately by paying the copyright owners, with artists able to opt out.
No product launch is believed to be imminent, and none of the companies mentioned would comment on the report when contacted by Sky News.
It comes as labels, artists, and fans alike grapple with the prominence of “deepfake music”, which has become increasingly convincing as generative AI tools become more powerful and accessible.
Sky News has reported on examples of a fake The Weeknd-Drake collab, a new cover by Michael Jackson, and Kanye West belting out some country tunes.
It’s led to calls for appropriate guardrails in the industry, though some artists are embracing it, with Canadian star Grimes already letting fans access her GrimesAI “voiceprint” to make songs that sound like her own.
Earlier this year, Universal Music said people will have to choose “which side of history” they want to be on – on the side of artists and fans or “deep fakes, fraud and denying artists their due compensation”.
The Grammys, the music awards show, has suggested the industry will need to adapt to the tech, with any work generated using AI to be made eligible for entry.
Concerns around AI are a hot topic across the creative industries, with fears it could be used to replace actors and writers a huge driver of strike action in Hollywood.