A song featuring the voices of Drake and The Weeknd called “Heart on My Sleeve” has amassed over 250,000 Spotify streams and 10 million views on TikTok. But the two renowned musicians had nothing to do with the song — an artist going by the name “Ghostwriter” generated the song using AI.
Drake and The Weeknd have not yet responded to the song, but Drake recently commented on AI-generated music that rips off his voice. When Drake noticed an AI model of himself singing “Munch” by Ice Spice, he wrote on his Instagram story, “This is the final straw AI.” It’s possible he was messing around, but he would be far from the first major artist to take issue with the rising count of deepfake songs.
In 2020, Jay-Z’s agency Roc Nation submitted copyright strikes against YouTube uploads of AI-generated Jay-Z deepfakes, but YouTube ended up reinstating the videos. And just last week, the same thing happened to Eminem; UMG, which represents both of these rappers, issued a copyright strike on AI-generated YouTube videos of Eminem rapping about cats.
Ghostwriter and Spotify did not immediately respond to TechCrunch’s requests for comment.
Copyright law is not technologically advanced enough to have specific guidelines regarding generative AI. But in the legal code’s existing state, transformative parody is permissible. However, these laws are very much open to interpretation, since the idea of what makes work “transformative” is subjective, and there is little case law to set precedent — historically, many of these cases have settled before reaching a judge.
UMG has recently taken steps to prevent the proliferation of AI-generated music that rips off its recording artists. According to a Financial Times report, UMG asked streaming services like Spotify to prevent AI companies from using its music to train their models.
“We have a moral and commercial responsibility to our artists to work to prevent the unauthorized use of their music and to stop platforms from ingesting content that violates the rights of artists and other creators,” a UMG representative said in a statement. The representative said that the rise of AI-generated music “begs the question as to which side of history all stakeholders in the music ecosystem want to be on: the side of artists, fans and human creative expression, or on the side of deep fakes, fraud and denying artists their due compensation.”
Once any kind of artistic work is part of a dataset, it can be hard to remove it. To help bring control back to artists, technologists Mat Dryhurst and Holly Herndon founded Spawning AI. One of their projects, “Have I Been Trained,” allows users to search for their artwork and see if it has been incorporated into an AI training set without their consent.
In some cases, though, removing one’s intellectual property from AI models can be like looking for a needle in a haystack. A living illustrator who has crafted detailed, high fantasy artwork for franchises like “Dungeons & Dragons,” Greg Rutkowski was one of Stable Diffusion’s most popular search terms when it launched in September, allowing users to easily replicate his distinctive style. Rutkowski never consented to his artwork being used to train the algorithm, and once the flood gates are opened, it might be too late for Rutkowski to regain the control he used to have over his work.
For now, Ghostwriter’s fake Drake and The Weeknd song remains on Spotify, but it may not be there for long.
AI music generators could be a boon for artists — but also problematic
Comment