When Universal Music Group CEO Sir Lucian Grainge says that successful artists should be compensated for the value they bring to streaming platforms, I agree with him. In fact, I think his argument points toward an even more fundamental reconsideration of the economics of music streaming that could get to the same end state.
During Universal Music Group’s October 29, 2026 earnings discussion, Mr. Grainge and UMG Chief Digital Officer Michael Nash raised important questions about how streaming services value music, particularly as platforms confront increasing volumes of AI-generated recordings. The underlying issue is whether a system that allocates royalties principally according to streaming volume adequately recognizes the contribution of artists who develop careers, build audiences, and attract paying subscribers.
That is a legitimate question, and Mr. Grainge is right to raise it. I would extend his argument one step further: rather than simply determining which artists deserve a larger share of an existing royalty pool, why not restore the ability of artists to negotiate the actual price of their music on a per-stream basis and forget the pool ideal altogether? And why not at least give them the freedom to withhold their recordings when a streaming service doesn’t want to pay that price?
The Problem With a Fixed Royalty Pool
The revenue-sharing model may have been understandable when subscription services were getting started. Streaming companies faced uncertain subscriber demand, competition from piracy, and substantial technological investment. Record companies wanted digital distribution to succeed. A fixed revenue share model offered a way to participate in its growth without imposing both potentially unlimited royalty liabilities and higher transaction costs from negotiation with individual artists.
But that was twenty years ago. Spotify’s third-quarter 2026 financial results reported 306 million premium subscribers, approximately €5 billion in quarterly revenue, and €672 million in quarterly operating income. We are no longer dealing with a struggling startup. Another way we know that is that Spotify co-founder Daniel Ek cashed out and is now going for the real money just like Google’s Eric Schmidt—blood money from AI drones and arms dealing.
Yet streaming still treats music royalties largely as an allocation of a negotiated percentage of service revenue. If a service has $100 available for royalty distributions, giving one artist an additional $10 generally means allocating $10 less to someone else—unless the service increases the amount available. This is the zero-sum problem of the Big Pool that is the only price discovery model in the world that treats all artists as of equal value.
It is not an unavoidable characteristic of music licensing. It is a consequence of the business model the parties chose. Ironically, the labels chose that model to help a startup that then turned around and bit them. The zero-sum problem becomes particularly acute when we try appropriately to recognize successful artists through preferential weights. Some artists contribute more economic value to a platform than others. Mr. Grainge is right about that. But why should their reward necessarily come from the relative shares of other artists which I doubt that he also wants?
Instead of saying, ‘Please give me a larger share of your royalty budget,’ an artist should be able to say, “Here is the price of my recording. If you don’t want to pay it, don’t play it.” That moves us from the Oliver Twist model of asking for a little more toward a genuine and orderly marketplace based on choice. Real democracy rather than the faux competition of the fixed pool determined by our tech overlords.
Free Agency Means the Freedom to Negotiate
In 1999, I wrote Why Free Agency Matters: The Coming Changes in Record Company Artist Relations. The paper examined how digital technology could change the relationship between recording artists and the businesses that distribute and commercialize their music. Technology could give artists greater freedom to select partners, negotiate terms, and choose how recordings reach listeners.
That did not mean record companies would become irrelevant. A successful and focused label can provide extraordinary value through financing, marketing, promotion, distribution, and career development. But I did mean that artists should have meaningful choices.
Twenty-seven years later, we have remarkably efficient global distribution but an increasingly concentrated practical relationship between musicians and the platforms that reach listeners, especially Spotify. An artist may have more ways to distribute recordings than ever before, yet less commercial leverage over the terms on which those recordings are consumed, at least on streaming platforms.
Spotify is so important to music distribution that artists and labels may feel unable to withhold recordings even when they believe a share of the Big Pool is inadequate. The ability to bargain is inseparable from the ability to decline an offer. Free agency ought to include freedom to choose when, where, and at what price a recording is available.
Two Royalty Systems Can Coexist
Consider two potentially available licensing paths. The first would be individually negotiated per-use licensing. An artist or label could negotiate a penny rate for qualifying streams, backed by terms that already exist today in streaming licenses—minimum guarantees, appropriate accounting, audit rights, and other commercial protections. The rate would reflect the recording’s value, the artist’s audience, competitive alternatives, and the parties’ willingness to make a deal.
This should not be reserved for major-label superstars. Independent artists and labels should have the same contractual freedom, even if their rates and bargaining leverage differ. An artist with low streams might begin their career in a collectivist Big Pool system and later negotiate directly after developing a commercially significant audience—a natural form of upstreaming, rather than an arbitrary algorithmic cutoff like the Discovery Mode free goods approach.
The second path would preserve collective royalty accounting for artists who prefer it. There are millions of recordings generating low to modest revenues, partly as a function of the Big Pool model. Separate negotiations for every track would be impractical. Pool accounting can provide administrative efficiency. Nothing about the proposal requires eliminating it.
But the systems must be separated correctly. Recordings on individually negotiated rates should not be counted in the Big Pool. Payments under those direct licenses must not simply be deducted from the royalties otherwise available to remaining pool participants. Otherwise smaller artists would subsidize the higher negotiated compensation due to the zero-sum problem. If one artist gets more, that reduces the per-stream rate for everyone else.
The pool for streams not covered by direct licences therefore requires its own defined and protected funding obligation, transparent calculations, and audit rights. Establishing that funding baseline is a substantive commercial negotiation, not simply a matter of excluding directly licensed streams from the denominator. The systems needed to administer both payment models already exist; the real challenges are economic and contractual.
The Ethical Pool: Artist Choice Within Collective Accounting
In 2018 I published Arithmetic on the Internet: The Ethical Pool Solution to Streaming Royalty Allocation. That proposal addressed a different, related question: who receives money inside a collective pool? It did not propose abandoning revenue sharing altogether. But that was almost ten years ago, and we have moved on to bigger problems that were predictable, but largely unaddressed until now.
Under conventional “streamshare” accounting, a subscriber’s royalty contribution is not necessarily directed only to the artists that subscriber listens to. My Ethical Pool (sometimes called “artist centric”) proposed that listeners could elect to direct their eligible royalty contribution to participating artists they actually played, while artists could choose whether to participate. The Ethical Pool and the conventional Big Pool would coexist.
That proposal concerned allocation, not the independently negotiated price of a use. A penny-rate system addresses the latter question. The two ideas are complementary: artists who prefer collective accounting could remain in a pool, possibly one offering an Ethical Pool option, while artists seeking their own per-use terms could pursue them. The connecting principle is choice.
AI Makes the Revenue-Pool Question More Urgent
Generative AI sharpens the problem with the royalty cap. Lucian Grainge has questioned what streaming-volume market share means when AI-generated tracks and recordings by working artists compete for listening. It is important to distinguish fraudulent streams, unauthorized uses of protected recordings or identities, and lawful listening to AI-generated music. Those raise different legal and economic questions. Given the abject failure of distributors and DSPs to deal with AI slop (see the University of Chicago study the Artist Rights Institute quoted in our comment to the Copyright Office’s consultation on streaming fraud), the royalty cap is in sharp focus today.
Spotify has said it excludes detected artificial streams from royalty calculations and maintains eligibility rules for certain recording payments. We should not assume that every AI track automatically earns royalties or that every questionable play dilutes the pool. Still, when eligible listening shifts toward new content without a proportionate increase in the revenue cap, all other recordings may receive smaller relative shares anyway. This is where we came in.
A negotiated penny rate would not eliminate AI competition or solve questions of copyright, consent, or rights of publicity. It would, however, separate the contractual compensation for a qualifying use from the changing volume of other eligible recordings competing for the same pool. That is one reason the structural distinction matters. (ARI argues that it would also put the economic incentive in the right place to combat AI slop and streaming fraud.)
Discovery Mode and the Cost of Algorithmic Dependence
Spotify’s Discovery Mode illustrates another aspect of the industry’s dependence on platform-controlled distribution. Eligible recordings can receive consideration in certain personalized recommendation contexts in exchange for a commission on the associated recording royalties.
Participation is optional, and an artist may reasonably believe the promotional opportunity is worth the cost. Nevertheless, the arrangement illustrates the imbalance created when a platform controls a significant part of discovery while also controlling the mechanics of royalty allocation. Too many artists spend time trying to please an algorithm rather than making and selling recordings. There are only so many hours in a day.
That is not, standing alone, proof of unlawful conduct. It is a reason to make sure artists have genuine commercial alternatives.
The Market Is Already Experimenting With Windowing
We need not speculate about whether artists would experiment with delaying streaming if they had greater freedom. Some already do. NxWorries released Why Lawd? on vinyl, CD, and cassette in June 2024 one week before making it available on streaming services. Jack White distributed physical copies of No Name before a wider digital release. Independent labels, including International Anthem, have experimented with physical and download sales before streaming availability.
In 2026 Sturgill Simpson, recording as Johnny Blue Skies, released Mutiny After Midnight on physical formats without an initial streaming or download release. Its reported chart performance showed that significant demand can be captured without immediate inclusion in the streaming catalog.
We do not know that these artists were intentionally trying to avoid release-parity obligations under streaming distribution agreements. There are excellent commercial reasons to sell physical recordings or downloads before streaming. But the practical effect can be similar: the artist establishes a window in which fans purchase the recording before it enters the streaming royalty system, without necessarily favoring one streaming service over another.
Bandcamp supports direct sales of recordings, downloads, and physical merchandise; Patreon supports memberships and early access. Neither platform proves that delaying streaming is always profitable. Both demonstrate that artists and fans value business relationships outside the pooled-subscription model. A $25 record sale is not equivalent to a $25 artist profit, but it is a direct transaction tied to a particular recording rather than a fractional claim on a service-wide royalty allocation.
The lesson is not that every artist should abandon streaming or return to CDs. It is that immediate streaming availability should be a commercial choice, not an unavoidable contractual condition of releasing music. The industry is already experimenting with release sequencing, sometimes through a cumbersome route. It should be possible to negotiate comparable flexibility directly in streaming licenses.
Bring Back Digital Windowing
There are prominent precedents. Taylor Swift initially withheld Red from Spotify in 2012, withheld 1989 in 2014, and removed her existing catalog. Adele’s 25, released in November 2015, was not available on major on-demand streaming services at launch and only reached them in June 2016. Those episodes do not prove that withholding is optimal for every artist. They establish that the choice can have commercial significance—just look to the bleating from Daniel Ek in mansplaining to Taylor and Adele why the earth was actually flat.
Suppose an artist proposes a negotiated penny rate for a new album and Spotify declines. Why shouldn’t the artist be able to license a limited exclusive window to Apple Music, Tidal, Deezer, or another willing distributor—or sell first through direct-to-fan channels—while leaving older catalog on Spotify? The artist could take a short-term risk in exchange for a better offer. A competitor could pay for valuable differentiation.
Some historical digital licensing arrangements have included broad availability, simultaneous-release, or parity commitments. But to the extent such terms prevent an artist or label from testing competing offers, they deserve reconsideration. Without the ability to withhold a new recording, the threat to walk away may be largely illusory. (Because what Spotify really wants is a compulsory license, so they are trying to contract their way to that end.)
A Competition Question Worth Asking
The contractual issue also raises a legitimate, fact-dependent antitrust question. A broad delivery obligation may have been commercially defensible while streaming services were struggling to establish viable catalogs. Its competitive consequences may be different in a mature market dominated in practical terms by a few large platforms.
A service’s scale or profitability alone does not prove monopoly power. Nor does every availability requirement violate antitrust law. Any analysis would need the actual contract language, a properly defined market, evidence of market power, the degree and duration of competitive foreclosure, and consideration of legitimate business justifications.
Still, a facially nonexclusive license may constrain competition if its accompanying terms deprive rival distributors of meaningful opportunities to bid for exclusive releases. That is a question worth investigating.
There is also a critical distinction between a label independently declining Spotify’s proposed terms and competing labels collectively agreeing to withhold music. The former is generally ordinary licensing discretion; the latter can create antitrust risks. The goal is independent competition for music rights, not a coordinated boycott.
Completing the Artist-Centric Argument
I think Lucian Grainge has opened the correct conversation. Successful artists create substantial value for streaming platforms, and the economics should recognize it. I would extend that principle by restoring contractual freedom rather than relying solely on a different distribution formula within the same fixed pool.
Allow artists to negotiate their own penny rates directly with streaming services. Spotify is the obstacle: it refuses to negotiate individual royalty rates or contract directly with artists, regardless of the price they might offer. Instead, Spotify requires them to use intermediary distributors, including its preferred distributor DistroKid, in which Spotify acquired an ownership interest. Artists should be free to negotiate independently, remain in the Big Pool if they choose, and explore more ethical, listener-directed royalty allocation. They should also be free to withhold or window new recordings when a streaming service refuses their terms—a big pool might be acceptable for back catalog but not for new releases.
These changes would not guarantee higher income for everyone. But they could encourage services to compete for valuable releases, give artists a genuine price-negotiation option, and avoid making smaller artists automatically finance every premium won by larger ones.
Twenty-seven years after Why Free Agency Matters, the same principle applies. Technology should increase an artist’s commercial freedom, not reduce it. Sir Lucian Grainge is right to insist that artists be compensated for the value they create. I would extend the argument to its logical conclusion: artist-centric streaming should mean the freedom to window, negotiate price, choose distribution arrangements, and say no when the deal isn’t good enough.
That is not a rejection of streaming. It is a proposal to make streaming work more like a competitive market which it definitely is not.
