Copyright Ambiguity: Streaming Services Remove AI-Created Drake-Inspired Song

A new AI-formed song featuring voices sounding eerily similar to the voices of popular musicians Drake and the Weeknd was reportedly yanked from numerous streaming platforms over alleged copyright concerns.

Blaze Media noted that while the singers are for now seemingly getting what they want with the takedown of the song, it remains unclear if they truly have the legal grounds to do such a thing.

The new hit, “Heart On My Sleeve,” was playable on various streaming platforms such as Spotify and Apple Music until Universal Music Group (UMG) requested its removal over claims of copyright violations, according to Tech Xplore. It had over 2.5 million “likes” on TikTok alone, along with over 600,000 listens on Spotify and an altogether total of over 15 million streams, reported Blaze Media.

“We will not hesitate to take steps to protect our rights and those of our artists,” UMG warned in the email, which BBC says was first obtained by the Financial Times.

Music Business Worldwide said that Apple Music, Deezer, TIDAL, and Spotify have all nixed the song from their platforms.

As was pointed out by Blaze Media, even the popular music website Genius, known for posting the lyrics of various songs, removed the piece’s script.

It goes without saying that many of these AI voice-mimicking technologies are arguably becoming so effective that they can be difficult to discern from a real person. Roberto Nickson, who reportedly is a designer working with Eluna AI, posted a video to Twitter showing just how similar AI-produced content can sound to that of a singer or rap artist.

“And just like that. The music industry is forever changed,” he wrote online. “I recorded a verse, and had a trained AI model of Kanye replace my vocals. The results will blow your mind. Utterly incredible.”

After playing the song he had just mixed, Nickson opined that this is “the worst” of AI we have seen and that in just a few years “every popular musician will have multiple trained models of them.” He also noted that user interfaces to use such technologies will become far easier to use.