mediagroup
capabilities
all services
music
publishing admin
sound
records
staff
new blog (4)

The Music Question That Never Changes: What Getting Licensing Wrong Actually Costs You


The Music Question That Never Changes: What Getting Licensing Wrong Actually Costs You

The question that never changes

Every music supervisor has heard some version of this request: “We are looking for something inspirational, like Coldplay.” Or exciting, like a Hans Zimmer score. Or whatever is #1 on Spotify this week. The names change, but the request does not. Clients want a feeling, not a band, and they always want it fast.

This request hasn’t changed in 30 or 40 years, and it’s not going to. What has changed is everything between that request and a finished, defensible piece of music. The tools have gotten faster, the terms are longer, and the paperwork is now denser. And somewhere in the gaps, almost nobody in the chain is actually checking where any of it came from.

The real protection was never the paperwork

I went looking for the friction I assumed was slowing people down and called a producer friend for his thoughts on making music in 2026 with the tools currently available. He shared that, when using sample libraries, as long as he has a subscription, he’s not downloading certificates of licensed use every time he creates with a sample. His subscription allows him to use what he wants. He also said he doesn’t know a single other working producer who operates any differently. When he delivers a finished track to a label or publisher, nobody asks him what tools he used to build it. There isn’t a checkpoint to answer to. Nobody is checking.

Television and film solved this problem a long time ago. What a TV/film producer has to hand a distributor before payment is not one document; it is a full legal deliverables package: chain-of-title documentation, title and clearance reports, music licenses, Errors and Omissions insurance with all the supporting paperwork, guild approvals, total credits documentation, and much more. Nothing airs or streams until somebody has assembled proof, in writing, of every right in the production.

A finished song delivered to a label or a publisher gets nothing close to that. At most: writer’s share, publishing splits, whether union musicians were involved, maybe a producer and engineer credit. Nobody hands over a title report. Nobody carries Errors and Omissions insurance on a single song. Even the one document that does travel with music into film and television, the cue sheet, only tracks how an existing song gets used in a specific program. It says nothing about how the song itself was made, whether a sample is buried in it, whether an AI tool touched any part of it, or who actually wrote it. Television and film built a complete legal standard for how a piece of music gets used. They never built one for where it came from.

There are real legal guardrails here. Copyright protects an original work automatically, the moment it’s created. But what you can actually recover depends on timing. Register before an infringement happens, or within three months of first publishing the work, and you’re eligible for statutory damages and attorney’s fees. Wait until after, and you’re generally limited to actual damages, which are often too small to justify a lawsuit at all. The law sets the guardrails. It does not enforce itself, and nobody at the point of delivery is required to show they are inside them.

The tools to close that gap already exist. Some examples: Splice can generate a certified license for a sample on request. Production music libraries register catalogs and document ownership splits. Delivery on the music side just does not require anyone to use them. Chain of title stays undocumented until somebody actually needs it, and by then there usually is not time to go build it.

The rule nobody wrote down

Here is what is actually keeping this from turning into a mess: professional judgment, applied one song at a time, by people who never write any of it down. My producer friend has a standing question he asks before he touches a new song he’s been asked to create. Did you use Suno? Is this a rough sketch of an idea, or did you run your own demo through it, or did you hand it your lyrics and prompt it to generate the song around them? Because he is going to stay as a work-for-hire. He’s not going to take composer or publisher credit. He makes whatever adjustments are actually going to improve the track. What he will not do, ever, is transcribe an AI-generated song and call the result his own original production. Once he delivers it, his job is done, and it’s now out of his hands. Any registration, release, everything downstream, becomes his client’s decision to make.

Then he made the statement that explains the industry’s whole relationship to this: “Everybody knows AI-generated songs get released constantly. Everybody also knows you never sync one. No music supervisor is ever going to place an AI-generated track in a film, a show, or a commercial.” And you know what? That rule is not written down anywhere. It does not need to be. It exists because people whose careers depend on defensible chain of title are standing in the middle of the process, and they already know exactly what happens later if they get it wrong.

What happens when nobody is in the room?

So, the real question is not whether producers and composers are cutting corners. They are not. They are applying judgment nobody ever asked them to formalize, and it’s working. The real question is what happens when an institution never gets access to that judgment at all.

A government media team or a university marketing department is not in the room when a producer asks the Suno question. They are not in the conversation where a supervisor quietly rules a track out for sync before a client ever sees it. When that same team goes looking for music on its own—a stock platform, a freelancer with no sync background, a tool that generates something in thirty seconds—they are not getting a weaker version of that professional screening. They are getting none of it, and they usually have no way of knowing that is what happened.

They are also the ones left reading consumer terms of service directly, with no professional standing between them and the fine print. If you write or review government contracts, you already know how this plays out. The “flow-down” language itself is generic; it just requires the prime contractor to pass certain obligations down to whichever subcontractors get hired. But somewhere in the same contract file, the actual companies get named. A subcontracting plan lists them. A consent-to-subcontract requirement means the government has to approve a specific one by name. The subcontract agreement itself names the parties in the signature block. The first generic clause is the trigger. Somebody, eventually, has to put a real name on paper.

Now take a case where the creator uses a royalty-free music option. The platform’s consumer terms have no equivalent to any of the government contracting workflows. The only place you would ever actually see the real list of third parties bound into that royalty-free music agreement is in discovery, after a dispute is already underway. Nobody wants to get there. Everybody just wants to use the music product and move on. And that is exactly the point.

The enforcement is already aiming at the pipes

Professional judgment works because it is personal. A producer or a supervisor has nothing to gain from getting that call wrong, and everything to lose. Once the same question stops living in one person’s judgment and turns into volume—millions of tracks, no one asking anything—it does not stay a private matter. It becomes a legal one.

None of this stays theoretical once there is no professional judgment anywhere in the chain, just volume. In September 2026, Universal Music Group sued DistroKid, the distributor that moves an estimated 40 percent of all new music released globally, alleging deceptive trade practices and a flood of AI-generated tracks presented as though they came from real artists. The initial complaint named 1,000 specific recordings, with a theoretical statutory maximum near 150 million dollars, and UMG called that figure the tip of the iceberg.

This was not the first time UMG had taken this posture toward a distributor. In November 2024, UMG sued Believe, parent company of TuneCore, alleging industrial-scale infringement: distributing songs sped up or remixed just enough to dodge automated detection, letting fake artist profiles under misspelled names hijack search traffic, and manipulating YouTube’s Content ID system to divert royalty payouts from the actual rightsholders. That case was settled in April 2026 and dismissed with prejudice, with terms undisclosed. This was a different situation from the DistroKid complaint. However, it showed a major label deciding that a distributor’s business model—not just an individual bad actor—was the problem.

None of this is the platform-liability theory the Supreme Court stated in the Cox ruling. These are claims about what the distributor itself did, not what its users did without its knowledge.

What is actually the target here? It’s not an individual artist. It’s not a single unlicensed clip. It’s the infrastructure itself, the pipe that volume runs through with nobody in the chain asking the question at all. Automated detection tools like Audible Magic and Pex already scan for exactly this kind of unverified volume, and they are only getting better at it every year. If a major label will send real legal resources to challenge the distributor sitting between creation and release, an institution operating three steps removed from any professional judgment is not in a safer position. It is in a less visible one, for now.

Adaptation is not the same as getting it right

Labels are attempting something like what my producer friend does one song at a time: asking the hard question before the work begins. But a label’s version of that question comes in a press release, not a private conversation, and a label has every reason to describe its own answer in the biggest light available. That is the difference between judgment and posture, and it is worth watching Warner’s Suno deal with that difference in mind.

I don’t think the answer here is to be wary of new tools. Technology changes what is possible, and the people who adapt to that early usually do better than the people who wait it out. Warner Music Group’s earlier settlement with Suno looked like a good example of that: a label getting ahead of a disruption instead of only suing it. It could have been seen as what a major label actually does when a disruption arrives.

Warner didn’t drop its lawsuit. They negotiated a structure: licensed models launching in 2026, with artists and songwriters in control over whether and how their names, voices, and compositions get used. That is a major label building documentation and consent into a disruption before the market forced the question, the same way a good producer asks the Suno question before he opens the session. That said, did songwriters actually have a seat at the table, or did the label decide on their behalf the way labels usually do? It’s not something the press coverage of their deal actually answers. A structure that is genuinely good for artists and a structure that is good for a label and described in artist-friendly language are not the same thing, and there is no way to tell which one this is from the outside.

What is not in question is how quickly the underlying claim ran into trouble. On September 18, 2026, Sony Music and Universal Music Group filed a second lawsuit against Suno, this one aimed at the new v6 model Suno built in partnership with Warner, BMG, and Believe. Sony and UMG argue v6 was trained in part on outputs from Suno’s earlier, allegedly infringing models, and that a new license does not erase what an older, unlicensed model already learned. Their complaint calls it the fruit of the same poisoned tree. Whatever Warner’s deal does for its own artists, it did not settle the question industry-wide, and it did not stop two of Warner’s own competitors from arguing that the whole model is still built on stolen ground.

That is the lesson here. Adaptation is not the same as getting it right, and a press release is not the same as a resolved question. Just as a line on a website’s FAQ page gets used as a marketing tool, it may not be reflected the same way in the Terms of Service Agreement. A company can move fast, wrap up the deal, and still be doing something that mainly protects the company. The industry has not agreed on whether these models are clean, and artists, like the institutions, are often the last ones told what an early deal actually means for them.

This was never about getting sued

I’d like to be clear. This is not a case against producers, composers, or the tools they use. The professionals in this business are already doing this right, asking the right questions, applying judgment nobody has to legislate. That part of the business is working.

Judgment like that has a boundary, and it matters to see exactly where it sits. I once had someone in business affairs put it to me plainly. He said, “We have a contract that tells us the rules of the game. We do not make decisions outside of the contract. What we do is follow what the contract tells us to do.” That is not cruelty; it is just what happens once you are dealing with an institution instead of a person. And if a dispute ever reaches a courtroom, professional judgment does not walk in with you either. Paper does. The terms you clicked "accept" on. The chain of title you can or cannot produce. A judge does not ask whether everyone in the business should have known better. A judge asks what is actually written down and whether you agreed to it. It’s not a defense to say, “Nobody actually reads the terms of an agreement.”

For now, most institutions using unlicensed music will never hear from a lawyer. That is not evidence the practice is sound and stable. It is evidence nobody has asked the question yet, in a market where enforcement is moving closer every year. When it does happen, it doesn’t take much. The Department of the Interior found that out this year, after a holiday social media post set to a Vince Guaraldi arrangement of a traditional song—the kind of gesture that feels harmless enough that nobody thinks to check the sync license first. The company that controls that Peanuts music catalog sued. Interior settled in August for $20,000, plus a new requirement to clear future use with the rights holder first. Nobody there set out to infringe anything. They just wanted something that felt like the holidays. Copyright exists to make sure the people who create music can actually earn a living from it. An institution that operates without the professionals who protect that, whether on purpose or simply without realizing they were ever part of the process, is making a bet against that principle without knowing it placed one.

The question worth asking instead

My suggestion is to set aside, “What does a music track cost?” That was always the wrong question to begin with. The better one: did anyone who would actually know to ask the right questions ever get the chance to?

At PrimalHouse Media Group, that is the question our catalog is built to have already answered, before anyone has to ask it.

Ryan Neill, CEO
PrimalHouse Media Group