On July 15, 2026, Tidal became the first major streaming platform to cut royalty payments on tracks where every creative component was generated by AI. The policy did not pull those tracks from the catalog. It stripped them of monetization and removed their eligibility for direct-to-fan sales, while making them visible to listeners with a badge marking their AI origin. Tracks that use AI to impersonate a specific real artist get removed entirely. Every other major platform, Spotify, Apple Music, YouTube Music, still pays on stream count regardless of how a track was produced.
That gap is the most important fact in music rights right now. Where a track lives, and how it was made, determines whether it earns anything.
What Tidal Actually Changed
The policy, reported by TechCom on June 29, 2026, draws a hard line at "fully AI generated." Under Tidal's definition, that means every creative element of the track, the composition, the arrangement, the performance, the sound design, produced entirely by a generative AI system without meaningful human creative contribution.
Tracks in that category now:
- Stay in the Tidal catalog
- Carry a visible AI badge on the listener-facing page
- Earn zero royalties
- Cannot be sold directly to fans through Tidal's platform
Tracks that use AI to simulate or impersonate a specific real artist's voice or identity are treated as a separate and more serious category. Those get removed, not badged.
The distinction matters. A track built in a DAW by a human producer who used an AI plugin to generate a chord progression and then re-harmonized it, recorded real drums over it, wrote and performed the top line, and mixed the result is not what Tidal is targeting. The target is a track where a user typed a prompt into a generative system, pressed generate, and uploaded the output.
Where the Other Platforms Stand
Spotify, Apple Music, and YouTube Music have not matched Tidal's move. As of the date of this article, all three platforms pay on stream count regardless of AI involvement in production, provided the track does not violate existing policies around impersonation, fraud, or manipulated streaming numbers.
Spotify has published guidelines on AI content that focus on transparency and fraud prevention, not on withholding payment from AI-assisted or fully AI-generated tracks. Apple Music has not published a separate AI-specific royalty policy. YouTube Music operates under the broader YouTube content ID and monetization framework, which addresses AI voice cloning and deepfakes under its existing impersonation rules but does not condition payment on the AI or human origin of a track's composition.
This means the same fully AI-generated track that earns nothing on Tidal can still generate royalties on Spotify, Apple Music, and YouTube Music today. That situation is unlikely to be permanent. Tidal moved first. Industry observers covering this shift, including Variety's reporting on the policy, have noted that the move creates a visible precedent other platforms will face pressure to address or explicitly reject.
What "Fully AI Generated" Means in Practice
The practical definition Tidal is working from has not been published as a granular technical spec, so understanding the edges requires reading the policy against what it is clearly designed to address.
Fully AI generated means a track where no human made a creative decision that was then executed by a human or a traditional digital tool. The generative system made the composition. The generative system made the arrangement. The generative system produced the audio. A human uploaded the file.
Contrast that with a working producer's actual workflow in 2026. AI tools are embedded in nearly every serious DAW and production environment. iZotope uses machine learning for mixing assistance. Melodyne's AI pitch correction has been standard for years. LANDR and similar platforms use AI in mastering. Splice's search and recommendation systems are AI-driven. Suno and Udio are the clearest examples of what Tidal's policy is actually targeting: full-track generation from a text prompt.
A producer who used Suno to generate a rough melodic idea, then re-recorded it with a live guitarist, layered in programmed drums, wrote and sung the vocal, and mixed and mastered the final track is in genuinely different territory than someone who generated a complete track and uploaded it. Tidal's policy is aimed at the latter. The former involves human creative contribution at multiple stages of the process that a human could document and defend.
The problem is that Tidal, and any platform that follows, will need a verification mechanism. Right now, the enforcement mechanism is not fully transparent. The badge system implies either self-disclosure by the uploader or a detection system, possibly both. Neither is foolproof. This is a real operational gap in the policy, and it will be tested.
What Artists and Labels Using AI in Production Should Be Doing Now
If you are making music that involves AI tools at any stage, start documenting your process. Not because you are doing something wrong, but because the policy environment is shifting fast and the difference between "human-assisted AI" and "fully AI generated" is going to be something platforms ask artists and distributors to certify or defend.
Practical documentation looks like:
Session files. Keep your DAW project files. A session file with MIDI data, audio recordings, plugin chains, and edit history is one of the clearest records that a human made decisions in a DAW. A track that was generated in full by a text-to-music system does not produce a session file. That absence will matter.
Version history. Save incremental versions. A track that started as an AI-generated rough and evolved through ten human editing sessions has a visible creative history. That history is your argument.
Sample and stem records. If you licensed samples, used a vocalist, brought in a live instrument, or commissioned any component of the track from a human, keep the receipts: invoices, contracts, credit agreements, recording session notes.
Prompt and output logs. If you used a generative tool for any part of the process, note what it produced and what you changed. This is not an admission of anything. It is documentation of your creative contribution relative to what the tool generated.
This documentation is not currently required by any major DSP for standard distribution. It is required by the logic of where these policies are heading.
The Label and Management Layer
For labels and management companies, the exposure here is portfolio-level. A roster with catalog distributed across all major platforms needs a clear picture of which releases were made with AI tools, to what extent, and how those releases are currently distributed.
If any of those releases are on Tidal and involve fully AI-generated components, they are currently earning nothing on that platform whether or not the label knows it. The badge system means listeners can see the designation. The royalty impact is immediate and already in effect.
For new releases, the due diligence question is now standard. When an artist or producer delivers a track, the production origin is material information for distribution decisions. Delivering a fully AI-generated track to a platform that has a royalty-cutoff policy for that category is not a rights issue in the traditional sense. It is a business decision with a predictable financial outcome.
Labels operating in the Latin, urbano, and regional Mexican markets face the same policy environment. Tidal has global distribution. The royalty cutoff is not geography-specific.
The Artist Impersonation Category Is a Separate Risk
Tidal's removal policy for tracks that impersonate real artists is more severe than the royalty cutoff and addresses a different problem. It connects directly to the voice cloning issue that has occupied the music industry since at least 2023, when AI-generated tracks mimicking Drake and The Weeknd circulated widely enough that Universal Music Group demanded platform action.
An AI-generated track that does not impersonate any real artist, even if the vocal is entirely synthetic, falls under the royalty-cutoff category. An AI-generated track that replicates a specific artist's voice, cadence, and signature sound crosses into the removal category.
For artists concerned about their own voice being used without permission, Tidal's removal policy is one response. The legal framework around voice likeness rights is still developing in the US and varies significantly by territory. The practical protection right now is platform-level enforcement, which is inconsistent across services, and direct registration with any AI voice protection tools that have emerged from the major label agreements reached in 2024 and 2025.
Why This Platform-by-Platform Gap Matters for Where You Distribute
The gap between Tidal's policy and Spotify's current policy creates a real distribution strategy question that did not exist eighteen months ago. A track that qualifies as fully AI generated under Tidal's definition earns nothing there and earns normally on Spotify. For artists or producers releasing in that category, Tidal's footprint in the overall royalty picture determines the actual financial impact.
Tidal's subscriber base is substantially smaller than Spotify's. MIDiA Research data cited in multiple 2025 industry reports puts Tidal well below 10 million subscribers globally compared to Spotify's 260-plus million. The immediate royalty loss from Tidal monetization cutoff is real but not catastrophic for most artists in terms of total royalty volume, assuming Spotify and Apple Music continue their current policies.
The larger risk is precedent velocity. If Apple Music or Spotify adopt comparable policies, the financial impact changes dramatically. Tidal moved on July 15, 2026. How Spotify, Apple Music, and YouTube Music respond in the following 12 to 18 months will determine whether Tidal's policy is an outlier or the beginning of an industry-wide standard.
What to Watch Next
Three things are worth tracking closely.
First, how distributors handle certification. DistroKid, TuneCore, CD Baby, and the other major independent distribution channels are the actual intake point for most of the AI-generated content reaching platforms. If Tidal requires distributors to flag or certify AI origin at the point of delivery, that creates an upstream chokepoint that shapes the whole market. No major distributor has published a certification requirement specific to Tidal's policy yet.
Second, whether detection systems become audible in the data. Tidal's badge system implies that some tracks are being identified without relying entirely on uploader self-disclosure. If a platform-level AI detection system is running, its false positive rate will matter enormously to artists using AI tools in legitimate hybrid workflows.
Third, legislative activity in the US. The NO FAKES Act and related federal proposals around AI-generated likeness and voice protection are still moving through Congress as of mid-2026. Their final form will affect the legal underpinning of removal policies across platforms in ways that go beyond what any single platform can decide on its own.
For now, the practical reality is this: document your process, know your platform's current policy, and watch distributor guidance closely. Tidal set the first marker. The rest of the map is still being drawn.