Tidal has released a formal policy on AI-generated music, dated 29 June, outlining how the platform will handle content that has expanded significantly across streaming services over the past year. The policy defines AI-generated music as work that is wholly or substantially produced by generative artificial intelligence, drawing a deliberate distinction between that category and the machine learning tools artists have routinely used in production for years.
The platform will continue to host AI-generated music, framing the decision as one of artist choice and listener autonomy. What changes is the level of scrutiny applied to that content. Tidal states that AI-generated material will be held to a higher standard of integrity than the rest of its catalogue, and it expects distributors to flag such content before it reaches the platform — an expectation the company says it will begin enforcing.
Starting mid-July, listeners will see an icon next to any track Tidal identifies as entirely AI-generated. As detection capabilities develop, the company intends to extend that label to releases it considers substantially, rather than wholly, machine-made. Alongside the transparency measures, Tidal is drawing a hard line on fraud. The platform will block or remove AI-generated music connected to impersonation, interference with genuine artists and their audiences, or bulk uploads paired with suspicious streaming activity. Tidal reserves the right to determine at its own discretion whether a release exists primarily to manipulate the system.
The most consequential element of the policy concerns royalties. From the date of publication, wholly AI-generated music will not earn royalties on Tidal. The company acknowledges that some AI-generated work — particularly material built on properly licensed models — may eventually warrant a share of revenue, but states clearly that its current priority is ensuring royalties flow to music written, performed, and produced by people. Until licensing arrangements between rightsholders and AI companies reach a settled form, Tidal says it will not knowingly pay out on tracks it considers entirely machine-generated.
The same rules apply to Tidal Upload, the service’s direct route for independent artists. AI-generated material uploaded through that pathway will be tagged and excluded from direct-to-fan monetisation, placing it under identical restrictions to the main catalogue.
Taken together, the policy stops short of an outright ban. AI music is permitted on the platform, labelled for listeners, and excluded from the royalty pool for the time being. The monetisation clause is the sharpest instrument in the framework — it places human authorship and machine output on different financial footing at the point where streaming income is distributed. How that boundary holds as detection tools improve, and as licensed AI models complicate definitions of what counts as AI-generated, will determine whether the policy remains workable as the broader industry continues to negotiate terms that do not yet exist in any agreed form.
Source: Decoded Magazine
