Last week, an AI-generated song featuring deepfakes of both Drake and The Weeknd’s voices went viral, though neither artists officially participated in its creation. Meanwhile, Grimes has suggested royalty sharing, offering 50% of her royalty earnings from any AI-generated song which uses her voice. The artist even claims she is interested in “killing copyright,” but this would ultimately undermine her ability to earn royalties from the AI-generated songs. While the music industry is now at an interesting juncture, musicians like YACHT and Holly Herndon have already started experimenting with AI technology as a creative tool to push boundaries in their music.
YACHT has trained AI on several years of its music and successfully synthesized the results into an album titled “Chain Tripping.” Similarly, Holly Herndon created a website called Holly+, which freely allows anyone to create deepfake music using her vocal cords.
The issue is, many artists are unaware of AI-generated music until it becomes too late. During a quarterly earnings call, Spotify CEO Daniel Ek commented on the momentous technology, while mentioning their policy of taking down the deepfake music of Bieber, Drake and the Weeknd. Despite the company’s stance, Ek was oddly optimistic, emphasizing that “more music” is always, in his view, “great culturally.”
For many musicians and music fans, there remains a deep unease about the use of AI-generated music. Entertainment lawyer Henderson Cole commented: “When artists are already struggling, it seems like a dangerous step,” as streaming payouts and the disruption of the live music industry have already left performers grappling for stability. Artists must protect their work from being depreciated, as AI technology has the potential to appropriate their likeness without consent.
Music has an intersectional influence in the development of new technologies. Many have already seen this with the crypto industry and their attempt to revolutionize the music streaming market. While some believe artists can relying on “fair use” legal argument as a shield against AI-generated music, Cole is less than enthusiastic; with Ed Sheeran and Robin Thicke both being sued for sounding similar to hit songs, he seems rather dubious of someone using AI to copy an artist’s voice and melodies. Currently, companies like Universal Music Group have been vocal about their disapproval in the use of generative AI. As we weave our way through this complex landscape of technology, the question remains unanswered – what is next for music?