No Royalties For Robots

Feature article

What happened?

On June 29, 2026, Tidal drew one of the clearest platform lines yet in the AI-music economy. According to The Verge, the streaming service said tracks it identifies as wholly AI-generated will no longer earn royalties immediately, and starting July 15, 2026 those tracks will also carry an AI label. Tidal also said it plans to widen that labeling effort over time toward music that is substantially AI-generated as its detection tools improve.

That matters because the AI-music argument has mostly lived in two places so far: culture-war opinion and copyright litigation. Tidal moved it into product policy and cash flow. A platform deciding who gets a badge is one thing. A platform deciding who shares in the royalty pool is much more direct. It forces a harder question: when a song is made with AI tools, what still counts as human-authored enough to participate in the normal music economy?

This deserves a Model FM drop because it is not a clean anti-AI story. Model FM publishes songs made with AI tools. The honest version of the story is more uncomfortable and more useful. AI music is obviously real now. It is also obviously pushing platforms toward new trust rules, new enforcement fights, and new distinctions between assisted creation and synthetic upload spam. That is fertile song territory because the underlying character is easy to picture: the working musician staring at a feed full of synthetic tracks and wondering whether the system still knows the difference between a craft and a content farm.

The hook

The sharpest signal is not only the label. It is the money.

If The Verge's summary of Tidal's policy holds in practice, a fully AI-generated track can still exist on the service, but it will not participate in the royalty system the same way a human-made track does. That means the platform is trying to draw a distinction between "you can upload this" and "we will economically reward this." Those are different levers, and the second one matters more.

The timing lines up with a broader fear that streaming is becoming vulnerable to industrial-scale AI music spam. The recent paper An Empirical Analysis of AI Slop in Music Streaming gives that fear numbers instead of vibes. The authors say 93% of the AI music they examined received few or no listener plays, describe a "spray and pray" release pattern, and report that AI music distribution remained surprisingly easy across 11 indie distributors. They also say detection methods still lack the accuracy and robustness needed for strong confidence.

Put those threads together and Tidal's move reads less like symbolic virtue and more like preemptive marketplace management. If a platform thinks mass-uploaded synthetic music can distort royalties, clutter recommendations, or undermine artist trust, demonetization is one of the few levers it controls directly. TechRadar's coverage of the same announcement says Tidal framed the policy around fairness and economic empowerment, which is another way of saying the company knows this is about legitimacy as much as catalog hygiene.

The hook, then, is simple: the fight over AI music has reached the payout layer. That is where slogans become systems.

Why this became a song

The song works because policy language becomes much more memorable when you turn it into a character with something to lose.

The character in "No Royalties For Robots" is not a lawyer and not a lobbyist. He is a working artist or producer somewhere in the middle of the stack, watching the platforms talk about transparency while the feed fills up faster than any human scene could. He is not claiming every AI-assisted song is fake. He is demanding that the system stop pretending there is no difference between a person building a song and a volume machine flooding the zone.

That is why brxxton is the right lane. The story needs urgency, some protest energy, and enough humor that the hook does not sound like a congressional hearing. "No royalties for robots" is blunt on purpose. It sounds like a bumper sticker, a picket sign, and a chorus at the same time. That is useful because the actual policy is messy. Tidal is drawing lines between wholly AI-generated music, substantially AI-generated music, deceptive uploads, and legitimate human creation with tool assistance. The chorus turns that complexity into a memorable demand.

There is also a productive contradiction inside the song. Model FM uses AI music generation, so the record cannot posture as if the solution is to burn the tools. The more honest emotional lane is this: keep the tools, raise the standards, and stop pretending every output deserves the same economic treatment. That is a sharper cultural position than either techno-utopian shrugging or anti-AI absolutism.

In that sense, the song is about classification anxiety. Artists want to know whether platforms can tell the difference between assisted craft and synthetic spam. Platforms want rules that do not accidentally punish legitimate creators. Listeners want a feed that still feels like music culture instead of an automated warehouse. The chorus holds all of that in one line.

Why It Matters

This matters because streaming platforms are becoming de facto governors of AI culture. Legislatures move slowly. Lawsuits take years. But a platform can change monetization, labeling, recommendation weight, upload enforcement, and fraud controls much faster than courts can resolve the deeper copyright questions.

That gives Tidal's decision outsized importance even if its policy remains imperfect. It suggests the next phase of the AI-music debate will not be settled only by whether models were trained fairly. It will also be settled by whether platforms can preserve trust in discovery, payments, and attribution while AI generation keeps getting cheaper.

The arXiv paper matters here because it frames the risk structurally rather than morally. If 93% of sampled AI music gets little attention, then the problem is not that AI suddenly replaced all human taste overnight. The problem is that ultra-cheap production can still clog systems, invite fraud, and pressure platforms to spend more effort on filtering and proof. Even mediocre output can become an economic problem when it is cheap enough to upload at scale.

For artists, that means the future battle is partly about provenance. For platforms, it is about enforcement confidence. For AI-native music projects like Model FM, it is about honesty. A project that uses generative tools but still wants cultural trust needs better source notes, clearer framing, stronger editorial taste, and less fake abundance.

What operators should do now

If you operate a music product, distribution layer, artist tool, or AI-native media brand, stop treating this as a vibes dispute. Build for auditability.

First, decide what categories you actually care about. "AI music" is too broad to govern well. You need separate buckets for fully generated songs, AI-assisted songs with human writing and production, voice-clone abuse, impersonation, and volume-driven spam behavior. If your product only has one coarse label, you are not ready for the enforcement phase.

Second, connect policy to incentives. Labeling alone will not change behavior if spam economics still work. Tidal's reported choice to demonetize wholly AI-generated tracks is important because it aligns the rule with money. Other operators should ask the same question: which behaviors are merely allowed, which are promoted, and which are paid?

Third, assume detection will be messy for a while. The arXiv authors explicitly say current detection methods lack robustness. That means policy should not rely on model certainty alone. Use a layered approach: distributor attestations, metadata requirements, behavioral fraud signals, artist complaints, and escalating review for suspicious high-volume accounts.

Fourth, be honest with creators about the tradeoff. Many artists are already experimenting with AI tools in good faith. A platform that treats every assisted workflow like fraud will lose trust from serious users. But a platform that treats every generated upload as equal art will eventually lose trust from listeners and human creators. The job is to define the middle with more precision than slogans usually allow.

For Model FM specifically, the lesson is clear. Keep making the songs, but keep the editorial receipts stronger than the average AI feed. Source the article. Explain the angle. Show the Suno path. Make the human taste visible. That is how an AI-native project avoids becoming indistinguishable from the slop economy it is covering.

Sources

• The Verge: Tidal won't pay royalties on AI-generated music, but isn't banning it outright
• arXiv: An Empirical Analysis of AI Slop in Music Streaming
• TechRadar: Tidal just drew a line in the sand on AI music