There actually isn’t a lot excuse for DistroKid to take a seat this one out. IFPI has launched its new Streaming Integrity Initiative, establishing baseline anti-fraud practices for music distributors. Whereas I recognize the hassle, I’ve to say that is fairly weak sauce in comparison with what I might do, however then I’m an Outdated Testomony form of man.
The IFPI commitments are hardly radical. Distributors are requested to know their clients (which is arguably their authorized obligation anyway), confirm rights, vet uploads for infringement, fraud and AI-related dangers, act in opposition to repeat offenders, share acceptable fraud intelligence, and measure whether or not their safeguards really work. The initiative goes additional in an necessary respect: distributors are particularly requested to determine AI-generated content material and potential infringements of voice, picture, title and likeness rights.
CD Child signed. So did FUGA, The Orchard, Symphonic, ADA, AWAL, Virgin Music Group and quite a few different distributors and trade organizations. DistroKid didn’t.
That omission deserves consideration as a result of DistroKid isn’t some marginal aggregator. It reportedly distributes 30% to 40% of all new music globally. And based on the dataset cited by Music Enterprise Worldwide, DistroKid has turn out to be a very necessary pipeline for commercially launched AI music: 75.8% of the AI tracks within the examine for which distributor data was accessible had been distributed by means of DistroKid, whereas 90.4% of the tracks studied had been generated utilizing Suno.
Then there may be Spotify. Spotify isn’t merely one other DSP receiving DistroKid deliveries. Spotify owns fairness in DistroKid. As lately as its June 2026 monetary statements, Spotify recognized its funding in DK Holdco LLC—DistroKid—as its most important long-term fairness funding.
That creates an uncommon chain of financial relationships: Suno generates the recording → DistroKid distributes it → Spotify streams it → Spotify calculates royalties → DistroKid participates economically as distributor → Spotify additionally holds an fairness curiosity in DistroKid. None of that proves misconduct. However it makes DistroKid’s refusal—at the very least up to now—to hitch a music trade initiative requiring primary AI, rights and fraud diligence significantly tougher to shrug off.
It additionally raises a extra basic query about Spotify’s revenue-share royalty mannequin. Spotify doesn’t pay a set royalty for every stream. Recording royalties are allotted based on a observe’s share of eligible streams throughout the fastened royalty pool. In different phrases, Spotify has a cap on how a lot they pay within the combination, however no consumer of Distrokid has a ground on how low Spotify can go. The extra eligible AI tracks Spotify can push out, the extra royalty-free tracks Spotify distributes and, if the tracks do entice a royalty, there’s an ever decrease share paid out.
In truth, Spotify itself explains that tracks should now cross a 1,000-stream annual threshold to take part within the recording royalty pool and that synthetic streams it detects don’t earn royalties. Which means an ineligible recording doesn’t merely obtain cash it shouldn’t obtain. Relying on how the pool is calculated, its participation can have an effect on the allocation amongst everybody else.
This turns into particularly necessary with absolutely AI-generated recordings that supposedly don’t get a share of the royalty pool. If a recording lacks enough human authorship to qualify for copyright safety, why ought to it take part in a royalty system essentially designed by statute to compensate human-created copyrighted works and performances? At minimal, DSPs and distributors ought to be capable of determine what they’re ingesting earlier than cash begins transferring.
And right here Suno V6 introduces a genuinely weird complication. V6 is now not merely a prompt-to-song generator. Suno describes a system that may edit parts of current user-created recordings and songs, mix a number of sources, isolate and pattern components, construct new beats round current materials, and protect components of an current recording whereas altering others.
That creates a doubtlessly necessary distinction between purely generated Suno output and a Suno-produced by-product incorporating substantial human-created materials equipped by means of a consumer’s account probably with out the consumer even figuring out it occurred.
Suppose a musician uploads an precise human efficiency—a vocal, guitar half, composition or accomplished recording—and Suno grabs that recording within the background and produces a brand new model incorporating substantial parts of the human materials. This might be permitted below Suno’s interpretation of its TOS. The ensuing recording might comprise sufficient protectable human authorship to current a really totally different copyright query from a track generated completely from a textual content immediate.
However whose human authorship is doing the work? It might be the Suno consumer’s. Suno’s phrases get hold of a very broad license to consumer submissions, together with rights to breed, modify, distribute and create by-product works from them. The license expressly reaches voice rights and different indicia of persona embodied in uploaded materials. On the similar time, Suno acknowledges that it can’t assure that copyright exists in AI output.
So we might arrive on the peculiar state of affairs wherein an AI-generated recording qualifies for downstream financial remedy as a result of Suno integrated sufficient human-created materials obtained from considered one of its customers which—in some way—finds its means into Spotify by means of a distributor like Distrokid. That ought to not turn out to be a copyright-laundering mechanism.
Nor ought to “some human materials is current” routinely reply whether or not a selected sound recording belongs in a royalty pool, who owns the related rights, whether or not the human contributor consented to the actual downstream exploitation, or who ought to obtain the ensuing cash. Suno itself form of acknowledges a few of these distinctions. Its present assist supplies say paid customers can commercially distribute qualifying downloaded tracks, whereas its remix guidelines impose separate restrictions relying upon who created the underlying materials.
All of which brings us again to DistroKid. The distributor sitting on the gateway is exactly the celebration able to asking these questions earlier than the recording reaches Spotify which is the route that IFPI goes (“Distributors, Properly Positioned to Forestall Uploads of Unlawful Fakes, Should Decide to Act In opposition to Streaming Fraud“). Who uploaded it? The place did it come from? Was AI concerned? What mannequin generated it? Does the uploader management the underlying recording and performances? Are there voice, title, likeness or different personhood rights implicated? Is the recording considerably human-made? Is it eligible for monetization below the AI service’s personal phrases?
These aren’t unique necessities invented by AI critics. They’re remarkably near the requirements the recording trade itself simply adopted. IFPI’s lately introduced rules for AI recordings likewise say chart-eligible AI recordings must be considerably human-made, generated utilizing correctly licensed providers, free from manipulation considerations, compliant with copyright and character rights, and appropriately recognized as involving AI.
DistroKid is among the largest gateways between AI era and the royalty pool. Spotify has an fairness curiosity in that gateway. And Spotify’s royalty system redistributes a finite pool based on eligible streamshare. In opposition to that background, “we haven’t signed but” isn’t a very satisfying reply. you, Spotify.
If CD Child can decide to figuring out its clients, checking rights, figuring out AI dangers and policing fraud, DistroKid can too. And Spotify ought to need it to.
