Universal and Sony versus Suno – who gets to own the future of music? The most important music process since Napster started. And everyone is listening.
Do you like Taylor Swift? Enjoy this while you can. Because on the horizon is the not so far-fetched possibility that at some point there will no longer be a need for a Taylor Swift – or a Drake, or a Sabrina Carpenter. One machine is enough. At the push of a button, she spits out a song that sounds suspiciously like everything she’s ever heard. And she’s heard a lot.
Anyone who doesn’t like that prospect is in good company — Universal and Sony, for example. A lawsuit is currently underway in a Massachusetts court that is shaping up to be one of the most important in the music industry in the last twenty years. One of the kind that you make into a film in retrospect – remember, back when Napster existed? Suno is the Napster of this decade: the moment when a technology demonstrates to the industry that its business model was based on an assumption that suddenly no longer applies. Only this time we’re not copying what already exists. This time something new is created – from everything that ever existed.
And that’s the problem. The question: How does the AI know what already existed? Who released the training data?
This is exactly where Universal and Sony come in. Because a machine that sounds like Taylor Swift heard Taylor Swift. The only thing that remained open until recently was whether this could be proven.
AI vs. AI: The proof is in the fingerprints
This began the march of the legal armada – equipped with experts, fingerprint technology and the patience that only comes when there is a lot of money at the end. To prove what Suno had heard, Universal and Sony first had to find their own recordings in the training data — and Suno wasn’t much help. So they used Audible Magic, an industry-standard fingerprinting technology that matches acoustic signatures like a forensic scientist matches papillary lines. AI against AI: In order to catch the machine that had heard everything, a second one that recognized everything was needed.
Expert witnesses traveled twice to a secure room at Suno’s law firm and digitally fingerprinted every audio file in the training inventory over two full weeks.
But it’s no longer about the bill for 61,026 songs. The question is whether the foundation of all AI music is even legal.
Fair use or breaking the law? Two exits, two worlds
Suno’s defense is “fair use.” The machine doesn’t copy anything, it just learns – and creates something new from it. The labels see it differently: Anyone who creates new songs from other people’s music that compete with the old ones for the same listeners cannot rely on mere learning.
If Suno wins, the dam is broken: then the training was legal, and every generator will take what they can hear in the future. The artist becomes the raw material for her own replacement – unpaid, unasked. If Suno loses, the opposite applies: training without a license becomes a violation of the law, and closed, licensed platforms become mandatory — for everyone.
A single ruling that changed the rules of the game for an entire industry for the next twenty years. Negotiations will take place in the summer. And while half the world is watching the World Cup in the USA, the music industry will be paying attention to this groundbreaking legal dispute.
