Putting aside the dystopic machines-vs-humans Terminator version of AI’s future, let’s look at the current landscape of AI and its positive and negative underlying factors by focusing on an actual business—the music industry.
Integrating artificial intelligence (AI) in the music industry has revolutionized various aspects of music creation, distribution, and consumption. Initially, AI technology was primarily used to create personalized playlists and recommendations on streaming platforms and automated music tagging.
However, with the advent of deep learning and neural networks, increased computational power, the availability of large and diverse datasets, and advancements in algorithmic techniques, the innovations made by generative AI in music have been exponential.
Still, AI’s full potential in the music industry is yet to be realized, and its recent rapid advancements have raised numerous legal challenges and implications. As AI advances, ongoing dialogue, collaboration, and proactive regulation will be necessary to shape an ethical future for generative AI that fosters innovation while guaranteeing fair compensation for “human” content creators.
Generative AI in Music: What’s Ethical, What’s Not
AI algorithms have become more sophisticated, enabling the generation of original compositions, lyrics, and even imitations of specific artists’ styles and voices. This year, someone using the pseudonym Ghostwriter977 put up the popular song “Heart on My Sleeve” but with fake vocals from two A-List musical artists—Drake and The Weeknd. Then Ghostwriter977 struck again, putting up a Bad Bunny and Rihanna AI collaboration. Both uploads were promptly met with take-down notices.
The response from Drake was clear and to the point when he wrote on Instagram: “This is the last straw.” His parent label, Universal Music Group (UMG), chimed in with this statement:
“The training of generative AI using our artists’ music (which represents both a breach of our agreements and a violation of copyright law), as well as the availability of infringing content created with generative AI on DSPs (Digital Service Providers), 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.” Source.
Following the UMG declaration, the industry publication Variety reported that Ghostwriter977, undeterred, had submitted both the Drake and Weeknd tracks “for Grammy consideration.” Then, in an interview with The New York Times, Recording Academy CEO Harvey Mason Jr. added another layer of controversy to the issue by declaring the Ghostwriter977 work “absolutely eligible” for a Grammy “because a human wrote it.”
But Mason quickly backtracked on his declaration. On September 7, 2023, he posted a video on Instagram in which he said, “I’m sorry, but I have to clear up some of this bad and really inaccurate information that’s starting to float around,” adding, “This version of ‘Heart on My Sleeve’ using the A.I. voice modeling that sounds like Drake and The Weeknd, it’s not eligible for Grammy consideration.” The Copyright Office has held that for a work to be copyrightable, it must owe its origin to a human being.
So, while history may not repeat itself, it certainly rhymes. Does anyone remember an infringer, Napster (Bad Kitty), that transformed into a legitimate content provider, Napster (Good Kitty)? The music industry already has a well-established method for licensing content. DSPs routinely license vast content from the major and independent record labels. This model could be used to train generative AI models while content creators are fairly compensated.
Some Artists Are Embracing AI
While Drake and UMG quickly crack down on AI-generated content, at least two artists welcome the AI era in music, with one setting her boundaries vis-a-vis compensation.
A singer, “Grimes,” who says, “I think it’s cool to be fused with a machine,” has also proactively set up a compensation model. Grimes expands, “I’ll split 50% royalties on any successful AI-generated song that uses my voice,” and further continues, “Same deal as with any artist I collaborate with. Feel free to use my voice without penalty.” Source of above quotes
Progressive music legend Peter Gabriel invites AI creators to collaborate with him. Partnering with Stability AI, the leading open-source AI company, Gabriel has announced the debut of a generative AI music video challenge. The #DiffuseTogether AI Automation Challenge is the first in a series of AI animation challenges to feature musicians of different genres.
The Use of AI in Music and Enhancing Human Creativity
The creative possibilities of employing AI are endless. For example:
1. Music creation and collaboration: AI-powered tools can aid musicians in providing inspiration for new musical ideas and assist in the creative process. Different genres, soundscapes, and combinations of instruments and styles can be blended. This can serve as a source of inspiration and facilitate collaboration between artists with diverse backgrounds.
2. Personalized music recommendation and discovery: AI algorithms can analyze vast amounts of user data, including listening habits, preferences, and contextual factors, to provide personalized recommendations to listeners. This not only benefits consumers by exposing them to new and relevant music but also helps musicians reach wider audiences.
3. Creation of new compositions from iconic performers: Deep learning AI-powered tools can generate new compositions from vast datasets of (properly licensed) music. The technology is only getting better. Imagine new albums from Elvis, the Beatles, Michael Jackson, etc. Artists (or their heirs) and music labels can generate substantial royalties with AI-assisted content creation.
Using AI to Ensure Fair Compensation
While AI does raise a host of ethical issues in the music business, it can also be an ally of music labels, artists, music publishers, and songwriters. For example, it can detect underpayments in royalty reporting from DSPs in very granular detail. It can determine underreporting by territory and ISRC number. In other words, AI helps rightsholders get paid correctly.
And AI is an ally to artists, labels, songwriters, and music publishers on another important front—revenue forecasting.
Using advanced algorithms, AI has the potential to make very accurate predictions in streaming counts among the different DSPs. Using AI to forecast streaming revenue can help labels make fair advance payments to artists. And it can help artists and songwriters ensure that their advance payments are fair.
To sum up, at this point, ensuring fair compensation for AI-assisted human creators presents a significant challenge, but the obstacles are not insurmountable. When AI is used to assist in the creation process, there needs to be a presumption of human creation to copyright the content. So, there needs to be a licensing model enabling copyrighted content to be used to train AI platforms. This model will ensure content creators are compensated fairly.
Finally, we can employ AI itself to make sure that happens. AI can detect missing streaming counts and underpayments in a way that was impossible a few years ago. Imagine getting to a place where artists and their representatives could have reasonable assurance that the royalty statements are complete and correct. That puts rightsholders and artists in a new and strong position. In short, using AI’s predictive algorithms should ensure complete and accurate reporting of royalties to all content creators.
Contact Joe Rust @ firstname.lastname@example.org if you have questions about AI and Royalty Audits.