The Music Industry Has an AI Problem

Given the rate at which Black culture is appropriated, it’s important to remain vigilant and call out blatant offenders. But because TikTok user Ghostwriter977’s identity and intentions remain elusive, dismissing “Heart on My Sleeve” as a clear-cut example of artificial intelligence colliding with racism and gross stereotypes — as we saw with the AI-generated rapper FN Meka — might be oversimplifying a complex situation.  

What is undeniable is that Ghostwriter977 robbed artists of their consent and likeness. Calling out theft isn’t just the responsibility of people engaging in internet discourse. Regulating the use of AI technology in music and legally holding developers and users accountable should be top of mind for everyone in the music industry — artists and executives alike.

Remarkably, the “Heart on My Sleeve” fiasco is happening alongside other copyright disputes, including allegations that Ed Sheeran ripped off Marvin Gaye’s 1973 classic “Let’s Get it On” for his 2014 hit “Thinking Out Loud.” While there are clear differences between the two cases, it’s a reminder that the music industry takes any semblance of theft very seriously.

In a comment under a TikTok post promoting “Heart on My Sleeve,” Ghostwriter977 wrote, “I was a ghostwriter for years and got paid close to nothing just for major labels to profit.” Ghostwriter977 could be a disgruntled amateur artist. They could very well be a troll whose motive is to devalue hip-hop and, to an extent, Black culture. Or they could be an agent of chaos trying to sow discord in the music industry. All of the unknowns make it imperative for the industry to act fast.

How Universal Music Group NV, home to both Drake and the Weeknd, responded after “Heart on My Sleeve” went viral may be an indication of how things could play out. As first reported by the Financial Times, the company sent a letter to streaming platforms asking them to block developers from using the label’s catalog to “train” AI technology.

The battle is in its infancy, but it’ll be up to the courts to decide what constitutes legal use of these platforms, who gets to profit and how much they can collect. 

For now, it’s in the best interest of UMG — which controls about one-third of the global music market — to step up, but who knows how long labels will actually be on the artists’ side or if this will become yet another way to exploit musicians. 

AI’s endless capabilities require ironclad copyright protections that go even further than the precedent set by Midler v. Ford Motor Co. In 1988, the Ninth Circuit Court of Appeals ruled that as a well-known entity, Bette Midler was entitled to control over her distinctive singing voice when used for profit. Clear laws and rights are needed on the books that protect artists and explicitly mention AI. 

Unchecked, it could be disastrous. “Heart on My Sleeve” is far from the first song to dabble in using AI, but it is one of the first to sound so convincing if you don’t pick it apart, piece by piece, bar by bar. Have you heard Drake’s recent records? With songs like “Way 2 Sexy” and “Search & Rescue,” Drake has been making music that sounds like AI Drake for years. “Heart on My Sleeve” could have passed for his latest single, and in its wake, more AI-generated Drake records have popped up, sounding virtually indistinguishable from the real artist’s work. 

Just three years ago, OpenAI put a song on Soundcloud that attempted to conjure a Beatles record. The result was something that sounded like cryptic organized noise that was both sanguine and malevolent. What’s happening now is that we are experiencing the first records that actually sound convincing, thus proving that the technology is adapting through progressive learning algorithms.

People’s eagerness to use AI technology at the expense of rap artists, as opposed to other genres, could easily be explained by the fact that hip-hop is the most popular genre of music among young people in America. As access to AI becomes more widespread and social media allows for anonymity, we must keep the conversation going around cultural appropriation and technology, knowing that the answers won’t always be as clear as black and white.

However, there is no gray area here: AI-generated music must be produced in a way that is equitable and fair for artists. That will require a substantial amount of vigilance. Get your good shoes on and tie them up tight, because there’s a long road ahead.

Bloomberg LP, the parent of Bloomberg News, also announced a large language model for finance last month.

More From Bloomberg Opinion:

ChatGPT’s Secret Weapon Is Artificial Emotional Intelligence: Parmy Olson

In the ChatGPT Age, Prompting Is the Language to Learn: David Lee 

AI Could Spell the End of Big Business: Tyler Cowen

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

H. Drew Blackburn is a writer and editor. He has written for the New York Times’s T Brand Studio, Netflix’s Tudum, Level, Texas Monthly, GQ, Okayplayer and Eater.

More stories like this are available on bloomberg.com/opinion