Two weeks back, Cherie Hu of Water & Music published a scoop on how Udio, a leading generative AI music platform, is approaching the development of its new AI application, Starstruck. A day later, Universal Music and Spotify announced a new licensing agreement for a similar AI-powered feature, allowing users to create fan-made covers and remixes directly on Spotify.

If you’ve been following along, you’ll know that one of the major debates in music over the past few years has been whether artists, labels, and publishers should license their work to generative AI platforms. But as licensing deals begin to move from speculation to practice, an important parallel question is how these AI platforms and features will allocate attribution, compensation, ownership, and control, and who decides on these questions in the first place.

Over the last year, for example, Udio has notched deals with Universal, Warner, Merlin, Kobalt, and Believe, covering two of three major labels and a wide swath of the ‘independent’ sector. In parallel, Warner has signed a deal with Suno, while Sony Music continues to litigate against both platforms, with Universal also onboard for the Suno suit. These platforms, meanwhile, continue to attract huge volumes of investment capital, all while seeking to keep their training practices opaque.1

While there has been plenty of critical theorizing about the announced licensing deals, there has been far less clarity about what they will actually look like in practice.

Cherie’s reporting is important because it offers one of the first public glimpses at what Udio’s licensed music strategy may look like in its realized form. I won’t rehash the full piece—you should read Cherie’s piece in full—but it clarifies several of the Starstruck app’s features and its underlying rights framework.

What I want to focus on here is how attribution, compensation, and ownership are structured inside the app, and where the value generated by these platforms and outputs ultimately flows.2

Who gets paid by generative AI platforms?

I’ve previously written about the political nature of attribution and compensation systems for generative music AI. The key point is that attribution isn’t simply a technical question. Attribution systems decide which forms of musical influence count, which don’t, and who gets paid as a result. In that sense, attribution is a political design choice. Different attribution systems will distribute value differently.

This brings us back to Starstruck. Two features are especially important for understanding how value will move through the platform.

The first is attribution and compensation for the generation itself. According to Udio CEO Andrew Sanchez, Starstruck’s attribution and compensation system is limited to the catalog that users explicitly invoke. When a user directly prompts or selects a specific artist, song, or recording inside Starstruck to generate an AI-produced musical output—for clarity, let’s just call this a “song” going forward—the platform identifies those artists and associated rightsholders, attributes the generated song to them, and routes compensation accordingly.

The problem, of course, is what gets excluded from this framework. As Cherie reports, during the private Starstruck demo session for which she received a recording, a Kobalt member asked specifically “whether Udio could identify when a generated track was stylistically or structurally influenced by specific copyrighted works in the core model’s training dataset, beyond the song or artist the user selected.” Sanchez’ response was essentially that Udio could not do this, while also suggesting that tracing those deeper, model-level influences is “fundamentally difficult, and perhaps impossible, given how current model architectures work.”

The result is that Starstruck appears to draw a hard line between two forms of musical influence. On one side is the artist, song, or recording explicitly selected by the user, whose influence is recognized, attributed, and compensated. On the other side is the potentially much wider set of copyrighted works that may have shaped the model through training. Those works contribute to the generation of songs by the model, but unless the user directly invokes them, they appear to be excluded from the platform’s attribution and compensation system.

The second issue is ownership. Sanchez explained that within Starstruck, the generated song is owned and controlled by the rightsholders connected to the original artist or recording the user selected.

In other words, the directly prompted artist and associated rightsholders aren’t only compensated for the generation itself. They also control the resulting recording and any additional value it produces inside Udio’s system—and potentially outside of it, too, depending on whether the platform’s walled garden actually holds (I have my doubts!).3

Taken together, these two features create a system where the majority of artists whose work helps shape the model may be excluded, while only the artists named in the prompt and associated rightsholders receive credit, compensation, ownership, and future value.

Music’s reverse-Robin Hood economy

An example helps clarify the issue.

Starstruck’s remix feature reportedly allows users to apply genre or style shifts to existing recordings. So let’s imagine that a user prompts Starstruck to generate a “reggae” remix of a Sabrina Carpenter song. Assuming Sabrina Carpenter’s music is included in the UMG catalog licensed to Udio, Sabrina and the rightsholders connected to the underlying song and recording would be compensated when the user directly invokes her work inside the app. Under Starstruck’s reported ownership model, those same rightsholders would also own or control the resulting AI-generated recording.

But what about the reggae part of the reggae remix?

If the model’s ability to produce something recognizable as “reggae” depends on having trained on reggae recordings, then the resulting song would need to be drawing value from many artists who were never directly prompted, but whose work was included in the model’s training dataset. Would an artist like Original Koffee, a very dope Sony/Columbia reggae artist, be compensated, even as Sony is currently suing Udio for alleged mass copyright infringement?4 Would a small local reggae band whose original music is available online and may have been scraped into a Udio training dataset get paid?

The answer appears to be no, unless those artists’ works were explicitly invoked by the user (or otherwise covered by some other licensing mechanism not yet made public).

Value is therefore routed only toward the prompted artist and their rightsholders, even as the model’s ability to generate the requested style depends on other artists, communities, and musical traditions.

The uncompensated artists aren’t only excluded from the direct revenue stream. They also face a marketplace increasingly crowded with AI-generated songs that imitate, recombine, or derive value from the styles and traditions they helped build. As these generated songs circulate inside Udio’s walled garden—and especially if they leak beyond it—they compete for attention with the very artists and communities whose work helped make them possible, further diluting the market for their music and their chance of building a sustainable career.

Altogether, this produces a reverse-Robin Hood dynamic (a phrase we seem to be hearing a lot with respect to major label driven ‘innovation’ in music). The largest rightsholders, whose artists are more likely to be searched, selected, remixed, covered, and licensed inside platforms like Udio, are able to capture the value from AI-generated outputs. But much of that value is produced on the backs of smaller artists, genre communities, and musical traditions whose works helped train the underlying model.

Value creation for whom?

Zooming back out, these issues should be seen as specific instantiations of broader political economy questions regarding power, control over investment, and who is positioned to capture value in music.

Warner Music’s COO Armin Zerza recently referred to AI as “the biggest value creation opportunity that this industry has seen for a long time”. From his major label positioning, and in light of Cherie’s reporting on Udio’s planned framework, this appears to be at least partially true. The question then is: value creation for whom?

As demonstrated above, Udio’s proposed licensing framework will produce poor outcomes for most artists, while creating significant upside for major artists, labels, publishers, and rightsholders—a category that increasingly includes institutional investment capital.

So how to fight back? From the perspective of artists and music workers, the immediate question is not only what a fair licensing system should look like, but who gets to decide on what is actually implemented in practice. Because artists currently have limited power to shape these frameworks directly, the fundamental issue is how artists can build the collective leverage needed to participate in future decision-making and ecosystem shaping.

Udio’s proposed licensing framework is a concrete example of the structural challenge I’ve written about previously, where artists, music workers, and independent music companies need to find ways to collectivize around a strategy for building power and leverage. You should read the full essay, but the core argument is that the people and organizations whose labour sustains music need more than better individual deals. They need collective structures capable of building and shifting power, and with them, the terms on which these systems are built.

That means developing bottom-up power, including values-aligned pools of capital—financial, social, and human—among artists and independent companies that can be directed to support alternative music infrastructure, including platforms, tools, rights systems, and distribution models that prioritize artist control (ownership and governance!), fair compensation, and collective benefit over value extraction. At the same time, it means aggressively pushing for policy, regulation, and enforcement across multiple fronts simultaneously, including competition policy, digital markets regulation, labour rights, copyright, and more.

The future is produced through power, and then path dependency. The decisions being made today about where investment flows and on what terms music is licensed will shape the conditions artists and music workers will navigate for decades into the future. It’s high time we organized for a seat at the table.

Some music!

Home Front is a really great band. When the camera clips to the keyboard player skankin’!🤘 LFG!

I’ve got an essay out today introducing Notes on Financialization, a new bi-monthly-ish roundup on MUSICx tracking major financial transactions across music. Think investment deals, catalog acquisitions, private equity exits, distro consolidation, debt instruments, share buybacks, and all the other fun (see also: depressing!) ways that capital keeps reorganizing the industry.

The series is directly inspired by David Turner’s indispensable, but now defunct, newsletter Penny Fractions, and its recurring feature “A Note of Financialization,” where David tracked music industry investment and consolidation with a critical eye. With deep appreciation to David for the inspiration, this new series is an attempt to carry a little bit of that spirit forward.

This first edition looks at a particularly bonkers stretch of music business activity, including Kobalt’s sale to Primary Wave/Brookfield, the BMG-Concord merger, Ninja Tune ending up inside a Bertelsmann-controlled company, major-label moves into distribution infrastructure, and the growing debt machine behind catalog acquisition.

The basic argument is that these deals shouldn’t be understood as disconnected bits of industry news. They are, when put together, clues that show us where power is moving, what kinds of companies are being built, which parts of the music value chain are becoming strategically important, and how the future conditions for artists, workers, labels, and listeners are being shaped.

Put more simply, the thesis is: investment power matters and we should make it legible!

Head over to MUSICx and give it a read! Subscribe! And, as always, let me know what you think!

Share

You should go read it! He makes a clear case that for artists to gain meaningful leverage and agency over the direction in which the music industry is heading, they need to come together in a way that allows them to exert a unified, strategic, and durable power. If, like me, you believe that democratic power and the agency to direct structural change is built through bottom-up organizing, Bill’s op-ed will be music to your ears.1

Of course, making this happen is complicated by a range of strategic and practical difficulties, not least the fact that the music industry operates globally while most relevant policies, and unions themselves, remain organized and regulated at the domestic level. As Bill notes, in some jurisdictions like the US, a move for artists to unionize must first be preceded by new domestic legislation that allows for this to be legal in the first place. The Protect Working Musicians Act, for example, was an attempt at introducing the right for artists to formally organize in the US, and hopefully will be reintroduced to Congress at some point.

Even within these limitations, we’ve seen the power artists’ voices can have in influencing and producing more equitable music industry structures. A recent example of this is the role the Future of Music Coalition, along with partners like the Artist Rights Alliance, played in pushing the Live Nation–Ticketmaster antitrust campaign into the end zone.

Regardless of the obstacles in the way, Bill’s core point is strong: Without formalized collective power artists will regularly remain on the sidelines while the industry is remade around them, often to their own detriment.

Building a values-aligned coalition in music

If you’ve been following along with my writing you already know that I feel strongly about the fundamental need for collective power building in music:

Ultimately, coordination and collective action are hard, especially at scale. Organizing is slow, demanding work. But if our analysis of these structural tendencies is correct ( and I believe it is), then there is no real alternative to it. If these dynamics systematically push music toward race-to-the-bottom outcomes, then individual moral choices or isolated acts of resistance will never be enough. The only adequate response is to build the forms of solidarity, organization, and collective power capable of contesting them. At the risk of sounding like a broken record, if we care about the future of music, we have to organize! - via ‘Wicked ‘Zero-Sum’ Game(s)’

Bill’s essay sits squarely within this line of thinking by laying out the basic logic of artist collectivization, which is to build power. I want to widen Bill’s argument slightly and suggest that artist power on its own is not enough. What we need is a broader coalition, anchored by working musicians and the independent sector, including artists, songwriters, independent labels, publishers, managers, and other music workers.

Artists, workers, and independents should be building values-aligned power from below, while also pushing from above to curb the oligopolistic dominance of major music companies and their private capital partners. If the independent sector is serious about shaping the future of music, these are the levels on which it has to organize. - via ‘Who Controls the Future of Music?

Artist-based leverage, especially through unionization, could generate significant power, particularly if artists were organized and disciplined enough to withhold their labour by pulling music from strategically chosen platforms and other key leverage points in the system. The recent artist-led campaign to remove music from Spotify offers a useful example of the kind of power a more organized approach could generate.

However, whether artists have the juice to go it alone and win against the massive and growing power of capital in music, including the major labels and private equity, is still an open question. As we’ve seen, simply getting into the room and having any say in the design of new industry structures is difficult. Independent artists’ and labels’ experience with the shift to so-called “artist-centric” payment systems on major streaming platforms like Spotify is a case in point. Despite the name, the model produces a reverse-Robin Hood effect in streaming payouts and was pushed through largely on Universal Music’s terms, with smaller players left to react rather than shape the outcome.

So, if the influence of capital in music is growing larger each day, then how might we build a strong enough countervailing power to challenge it and push for a more democratic, equitable, artist-friendly, and humanistic music ecosystem?

Let me briefly acknowledge that what I’m suggesting may seem downright utopian to some given the music industry’s current fragmentation and the ever-looming power of the largest music companies, big tech platforms, and private capital. But that something appears difficult shouldn’t stop us from taking it seriously, or from imagining practical paths toward it. I often return to Mark Fisher on this point, whose work reminds us that there is real political power in the simple act of imagining different futures. Our politics, thinking, and building should be prefigurative. We can build real utopias.

So I’ve started to imagine what a future might look like in which working musicians and the independent sector build power together. This requires thinking through the shared values, strategic alignment and sequence of steps, issues, and focus needed to build a durable coalition of music stakeholders with enough collective leverage, and eventually pooled, values-aligned capital, to win meaningful change in the music industry.

This project is very much a work in progress. It has moved slowly (no, this work doesn’t pay my rent!), but I think it’s still worth sharing at this stage as a proof of concept that better futures can be imagined and pursued in practical ways.

Mapping power and values

I’m working across two approaches. The first involves mapping music industry stakeholders in relation to major ongoing policy and structural issues in music, starting with the US case. This is a form of power mapping designed to identify where stakeholder positions and incentives already overlap.2 By tracing those points of alignment, we can begin to see more clearly where a potential coalition already has common ground, and where it may be possible to begin building trust, relationships, and a strategically optimal sequence of shared priorities and issues on which to focus.

This approach also helps reveal where alignment doesn’t yet exist. In those cases, the mapping can surface the trade-offs and tensions that need to be worked through in order to build a shared position and stable coalition. As I’ve written:

Leveraging this kind of power would require coalition partners to make tradeoffs at different times. For that to work, participants in the coalition need to see one another not as competing interests with occasional overlap (i.e. one-off policy-issue dance partners🕺), but as part of a common collective with a shared set of stakes. We might even call this a kind of “class” formation within music, with a co-developed, material, values-based politics from which to act and organize.

Below is a screenshot of what this looks like so far. It’s very much a work in progress, with stakeholders, positions, and issues all in flux as I dig in on the research. Still, I think it gives an indication of where the project is headed and what it could look like in a more complete form.

My second approach comes at it from the other direction, and focuses on what alignment around core values might look like across this imagined coalition. If power mapping helps with the practical side of organizing, this second approach is about developing a shared normative framework: a sense of how the music industry ‘ought’ to be structured and governed, grounded in common values. I’ve written about this before in calling for a new politics of (independent) music rooted in clear commitments. This is the starting point. For a coalition like the one we are imagining to actually build and wield collective power, it needs to be grounded in a coherent politics of shared values and mutual accountability that can also guide stakeholders’ practical decision-making.

As of Secretly Group points out, a central challenge in developing this type of coalition is defining values at the right level of specificity. They need to be concrete enough to make accountability meaningful in practice, which means clear, legible commitments that can guide the choices of artists, who might be deciding which label(s) to work with, and fans whose music consumption might be swayed by the ethical commitments of artists and their labels. They also need to be clear enough to make some form of internal accountability within the coalition possible. At the same time, they need to be broad enough to make coalition-building realistic and to allow for the definitional ambiguity that, as Darius points out, is inherent in the ideas and history of independence in music. To not smother the place where the creative magic happens.

Likewise, as I have argued, this shared values framework must emerge from the bottom up, shaped through broad input from each stakeholder group involved and their individual members, including artists.

The point is not to arrive immediately at a finished doctrine or to impose a fixed set of values from above. It is to be more honest about the fact that the debates we’re already having about music are, at core, debates about values. They are about how the music ecosystem ‘ought’ to be and be built together (the ‘normative’), even if we’re talking about practical things like partnerships, who buys which companies, and what that means in terms of outcomes for artists and the music ecosystem.

Admittedly, this is a lot to balance. If we’re to get anywhere together, it might help to have a starting point for discussion and debate. In that spirit, I’ve begun pulling together some rough ideas about core values that might animate such a coalition. These ideas draw broadly from a range of traditions and histories within music, especially the independent music community, as well as from the stated positions of the stakeholders this proposed coalition would bring together, including artist groups, and label and publishing trade bodies. They’re also shaped by a wide range of public and academic thinking across music business research, political economy, cultural policy, labour, regulation, and more, along with my own experience in music and artistic communities over the years.

Here’s where my thinking has landed so far. Importantly, it begins from a core commitment to the idea that solidarity can produce better outcomes for everyone over the long term, even when it requires giving up perceived short-term advantage:

These are, of course, only my own initial thoughts. I’m putting them forward in the hope of sparking broader discussion and building momentum.

Together with the power-mapping approach outlined above, they’re intended as a starting point for the difficult, shared work of articulating a common values framework, defining what it means in practice, and building the political strategy needed to advance an agenda grounded in it.

In that spirit, I’d love to hear comments, additions, counter-proposals, wild left-field ideas(!!), and anything else that could help push this discussion forward. See you in the comments 🤘!

Some music!

One of the prouder moments from my grant-writing days was helping Joseph secure funding to make this sick rescore of Toy Machine’s legendary 1996 skate video Welcome to Hell. Jos even rescored a bunch of the foley sounds of skateboarding, as the original clean audio from the film was lost to history. Très cool!

Spring is here in Ottawa, and I live just around the corner from the skate park. I love watching everyone rip and listening to this album in the ol’ headphones 🤘🛹

I’m happy with that essay. But it is dense!

The central argument is that practices like trend simulation are part of a broader class of race-to-the-bottom collective action problems, and that addressing them requires a diverse range of approaches. One way to think about this is as a toolkit. If we want to solve the problem, we should probably be using every tool available.

And so, I think it’s worthwhile to try and visually clarify the macro-approaches available. Broadly, they look like this:

As soon as I made this simple diagram my completist brain thought: You should turn this into a full typology with more detailed explanations! And so I did that…

You can view it in full HERE.

My hope is that this visual approach makes the overall framework more intuitive and accessible. What I find particularly useful is how it makes it much easier to see how responsibility, both moral and practical, is actually spread across almost all of the stakeholders involved. We don’t have to choose which group to point blame at, but neither can we throw up our hands and ask, “Won’t somebody do something?” while exempting ourselves. Responsibility is shared, and so is the obligation to work to make things better!1

Some music!

A little Monday momentum 🤘🎚️🔊

Since then, a Billboard interview with Chaotic Good, a marketing firm that uses clipping and related fake marketing tactics—what they bleakly term ‘trend simulation’—has kept the topic alive. Much of the online discourse has focused on whether artists should use these types of strategies, and what it means if they do, both for themselves and their fans.

Much less attention, however, has been paid to the question of what might actually help address the problem, whether through artist-led responses or through policy and regulatory tools that could at least meaningfully disincentivize it. In part, this is because the conversation keeps getting stuck at the level of individual moral rationalization. In the Billboard interview, one of the marketers in question defends his choices by arguing that “Unfortunately, a lot of the internet is manipulation.” This is a familiar ethical sleight of hand, where the prevalence of wrongdoing is recast as permission to join in.12

However, artists, audiences, and policymakers do not have to simply accept wonky unethical marketing as a fixed condition of digital life. The task, then, is to identify the approaches and interventions that could begin to make these practices harder to justify and profit from.

As I have written, the problem isn’t simply that these types of strategies are ethically suspect on their own terms. They’re also a symptom of a deeper structural condition that is the zero-sum attention economy, and the ways in which it rewards short-term, individualistic thinking and behaviour, even as it degrades the broader conditions in which music circulates, connects, and generates meaning.

In that sense, trend simulation is part of a wider race to the bottom, and a collective action problem that can only really be fully overcome through coordination (we’ll come back to this later). But this structural tendency doesn’t mean we’re totally bonked. These are, after all, just markets. Humans make them, and if they aren’t doing what we want them to do, we also have the power to reshape them. So where can we intervene to make the game a little more fair?

A quick but important aside for readers: While I’m grounding this discussion in the Canadian context, as it is where I live and am most interested in influencing policy, my point isn’t just to produce a uniquely Canadian policy wishlist. It’s also to sketch out the broader menu of regulatory tools that could travel across jurisdictions. Canada is simply my main point of reference. I also refer to existing regulatory approaches, like the EU’s DSA, throughout this essay. While the specific language, design, and implementation of such interventions will vary by context, the underlying problems and approaches to tackle them are generally transferable. What follows, then, should be read both as a possible Canadian reform agenda and as a portable menu of options for regulating tactics like trend simulation.

Trend simulation as deceptive marketing

At the basic level, trend simulation should be understood as a form of deceptive marketing. When a song is promoted through paid or incentivized posts that are made to look like spontaneous cultural uptake, the commercial nature of that promotion is being concealed.

From that starting point, the seemingly obvious solution is simply to ban the practice through enforcement against the individuals and organizations engaging in it. In practice, that is easier said than done.

Trend simulation describes a broad landscape of fast-moving digital marketing tactics—one that sits adjacent to online scams—that often emerge more quickly than policymakers and regulators can assess or govern them. This delta between the velocity at which new practices emerge and the slower pace of regulatory response can create a space for grey-hat marketing, where practices like clipping, that are at least adjacent to—if not pretty clearly in—unethical territory, are still able to take hold because they are not explicitly prohibited and/or regulators lack the capacity to effectively enforce the rules already on the books. The problem may also be worsening owing to the knack that LLMs possess for generating code and undertaking programmatic tasks, making them particularly useful for things like developing and deploying marketing campaigns at scale.

In Canada, some of the legal and regulatory architecture needed to address this category of harms already exists, though it remains incomplete and could be strengthened. The Competition Bureau has provided guidance that the deceptive marketing provisions of the Competition Act apply to influencers and require disclosure of “material connections”.3 Under the Act, courts can order a person to stop the conduct, publish a corrective notice, and pay administrative monetary penalties.45

Canada could go further here. The U.S. Federal Trade Commission, for e.g., has moved beyond simple disclosure requirements with a rule that explicitly prohibits buying or selling fake indicators of social media influence. But again, a rule is only as strong as its enforcement, and the administration of the FTC has changed since the rule was introduced under Lina Khan’s FTC.6 Nonetheless, a stronger Canadian version of this policy agenda would take the US tactic and clearly ban fake social proof and hidden paid cultural promotion, rather than relying only on general deceptive marketing principles.

The problem is also that existing tools haven’t been deployed systematically and more importantly, that they face real limits when confronted with the scale of the problem. While there is some precedent in Canada for using the Competition Act against fake forms of commercial social proof7, enforcement can’t realistically proceed one deceptive post, campaign, or intermediary at a time. This is especially true when (1) regulators depend on ongoing cooperation from platforms to identify deceptive marketing and access the data needed to investigate it8, and (2) when enforcement capacity is itself being reduced, as in Canada, where the Competition Bureau recently slashed 24 jobs (representing ~5% of its staff).9

Regulating the platforms

So, if trying to regulate every single use of trend simulation at source is a net with a giant lack-of-capacity shaped hole in it, what other tools might help us contain the problem?

As I have noted, the issue is structural and shaped by the platforms that host, rank, and amplify this type of marketing content. This is part of the justification for regulating platforms themselves, not just individual unethical marketers. Effectively, this means internalizing more of the cost and responsibility of enforcement to the platforms that are structurally positioned to detect, govern, and limit these practices at scale.

This narrows the enforcement burden on regulators (effectively increasing what we might think of as their enforcement productivity). Instead of chasing every individual instance of trend simulation, they can force platforms to take responsibility for detecting and limiting it, and then hold them accountable for whether those systems are actually working. This is appropriate because platforms aren’t neutral intermediaries. They optimize for attention and time spent on platform, and in turn profit from the audience dynamics that trend simulation amplifies (e.g., infinite scrolling). This creates a structural contradiction in platform incentives, where they are financially motivated to tolerate trend simulation, and consequently have weak incentives to enforce their own terms of service.

Some platforms do formally prohibit aspects of trend simulation type behaviour in their terms. For example, TikTok indicates that they bar fake engagement, fake personas, impersonation, and platform manipulation, while YouTube prohibits fake engagement and spam or deceptive practices. But the apparent scale of trend simulation reflected in recent public discourse suggests that these rules are either too weak, insufficiently enforced, or both. Recent reporting on Meta, for instance, suggests that fraudulent ads have continued to run on their platforms despite violating existing policies, while stronger advertiser-verification measures were reportedly shelved to avoid disrupting ad revenue. This gap between rule-making and policy on paper, and the failure to curb these issues in practice only underscores the contradictions and limits of platform self-governance.

What’s needed is democratically backed regulation and enforcement designed to change the incentive structure under which platforms operate, so that deceptive promotion becomes a cost rather than a profit centre.

More specifically, the goal of this platform-focused regulatory approach is fourfold: (1) to make hidden promotion more transparent; (2) to require platforms to actively govern hidden promotion through stronger terms, systems, and enforcement, while holding users accountable under deceptive marketing rules; (3) to bring the black box of social recommendation and discoverability into view; and ultimately, (4) to hold platforms accountable for the systems they design and operate, and for the social and economic outcomes those systems produce.

The EU’s Digital Services Act (DSA) provides a real-world model from which Canada can learn (and straight up borrow). The Act requires greater transparency and user control around ranking and recommendations, requires ads to be labeled, and obliges very large platforms to maintain ad repositories. Those duties are backed by enforcement powers, where regulators can demand information, inspect systems, order platforms to change their practices, and impose major fines for non-compliance, including penalties of up to 6% of a company’s global annual turnover. These powers have already produced results. In 2025, the Commission found TikTok’s ad repository wanting and forced binding changes requiring fuller ad disclosure, quicker updates, and better searchability. The European Commission has also shown it’s willing to escalate. X was fined €120 million under the DSA, including for transparency failures in its user-verification system and advertising repository.10

In the UK, the Competition and Markets Authority (CMA) has similarly pushed platforms to take more proactive steps against hidden advertising, including through developing compliance principles. Following a CMA investigation, Instagram was made to introduce clearer platform rules, better disclosure tools for paid promotion, and stronger systems to identify and act on undisclosed sponsored content.11

The stick of state-backed penalties, it would seem, can be a powerful tool. The apparent success of that approach is reflected in the UK’s decision to further strengthen the CMA’s powers in its 2024 passage of the Digital Markets, Competition and Consumers Act (DMCCA), giving it greater authority to compel information, order firms to change or stop illegal practices, and impose fines of up to 10% of global turnover for breaches of consumer law.

The current Canadian toolkit in this area is piecemeal, though Parliament is slowly moving towards acting on the harms produced by platforms. The most immediately usable federal tool is still the deceptive-marketing law approach discussed above, but it suffers from the enforcement capacity issues already noted. And even where it can be used, it is still a system for responding to deceptive conduct after it has happened (ex-post), rather than an ex ante regime requiring platforms to shape themselves in ways that structurally change the incentives for undertaking deceptive marketing in the first place.

One positive is that Canada has already accepted that digital platforms can be made to serve public goals. We see this in recent legislation directed at platforms. The Online News Act (Bill C-18) creates a framework aimed at forcing dominant platforms to compensate news businesses. The Online Streaming Act (Bill C-11), meanwhile, gives the CRTC authority to impose obligations on streaming services, including a requirement that many contribute 5% of their Canadian revenues to the broadcasting system, which also supports Canadian artists. In both cases, platforms have had to be dragged into their new regulatory boundaries kicking, screaming, and threatening.12

While there has been meaningful movement on platform regulation in recent years, Canada still lacks a broad regulatory regime for social platforms, including their advertising and recommendation systems. The closest it has come is Bill C-63, the proposed Online Harms Act.13 Though aimed at different harms, the bill drew on similar regulatory logics as those at issue here, including platform duties, record-keeping requirements, researcher access to platform data, inspections, compliance orders, and significant administrative monetary penalties.

What’s on the regulatory menu?

The next question, then, is what a coherent regulatory regime aimed at curbing harms like trend simulation and fake advertising should require of platforms.

To start, Canada should move beyond its current patchwork of deceptive-marketing law, regulatory guidance, and voluntary platform practices by imposing clear statutory duties on large social media platforms (what the EU calls VLOPs, or very large online platforms), drawing on the regulatory model the EU has already begun to establish. For the purposes of regulating trend simulation, three types of duties seem especially relevant.

First are basic visibility obligations including clearly labeling paid commercial content, maintaining public repositories of paid promotional campaigns, preserving historical versions of ads, and just generally providing more algorithmic transparency and control to platform users. The regulation of algorithms has rightfully become an issue across other domains in Canadian society, and is one with political momentum behind it, so is ripe for action.14 Second are identity and verification obligations, like advertiser verification and basic know-your-customer (KYC) rules. Together, these approaches wouldn’t eliminate deceptive promotion outright, but would at least raise the cost of impersonation, make throwaway advertiser accounts harder to cycle through, and reduce the information asymmetry in which platforms know vastly more than anyone else about who is buying reach and how they behave.

Third are oversight and accountability obligations, including requiring platforms to report on fake engagement, repeat advertiser abuse, and coordinated inauthentic promotion, while giving researchers and watchdogs meaningful access to platform data and independent means to scrutinize how platform systems are actually working. Strategies like trend simulation often operate at scale and are only truly visible when viewed as patterns spread across many accounts. Independent research, backed by meaningful access to platform data, can help identify these cross-account patterns and develop more effective ways of detecting and disrupting them.15

None of this is wildly alien to Canadian policy thinking, either, with proposals along these lines put forward by policy institutes like the Canadian Centre for Policy Alternatives and the Canadian Anti-Monopoly Project. Likewise, an agenda focused on cracking down on big tech platforms in a way that benefits artists might also be particularly resonant for, say, an insurgent left party with a new leader.16

Bottom-up is better

Of course, measures like those described above wouldn’t eliminate trend simulation and scammy marketing practices altogether, but they would make them harder to disguise, easier to detect, and less effective at scale.

So, if we can reduce the degrees of freedom for trend simulators to hawk their wares by regulating at the platform level (from the top-down), can we also ask artists, labels and other industry stakeholders to play a meaningful role from the bottom-up to disincentivize these practices?

My starting point for this would be to point to how the advantages of collective action and simply working together at any scale can mimic the effects desired from trend simulation campaigns. If the goal of these campaigns is to reach as many potential fans as possible via extended socials placement, then why not invert the logic and simulate trend simulation by doing the real thing. See what I did there? Recent music industry analysis alongside years and years of DIY music history tell us that the best pathway for emerging artists to overcome zero-sum, race-to-the bottom attention logics is through building careers based in community, focused on steady sustainable growth and building authentic connection.

This looks like loose collectives of artists who regularly hang out and create music and meaning together, tour together, share administrative work, and also, cross-post, share audiences, appear in one another’s content, and generally treat promotion as a shared rather than purely individual task. Yes, what I am describing here looks a lot like a values-aligned collective or independent record label, or at least a label-like formation. This is kind of the point across a lot of my writing. We don’t need to reinvent the wheel or the label or the distributor, and so on, but rather need to rededicate ourselves to living our values in practice in how we organize and operate. There is no technology that is going to get us out of the fundamental need to cooperate with and care for each other. Not to mention that the technologies that do actually work best, like cooperative and public governance structures, lean into those needs, not away from them. One way out of the zero-sum attention trap, then, is to (re)build some of these cooperative structures that make it possible to generate reach and cultural momentum collectively.

The nice thing is that doing this is arguably easier than ever. It can happen locally, with music-loving friends, and the bands you meet on tour. Club Records is a great example of this in Ottawa. But it can also happen across distance, through online affinities and deliberately cultivated networks of artists who actually like one another’s work and want to help it travel. Collaborative creation and release experiments like Wild Awake and Songcamp’s ‘headless band’ Chaos come to mind here. In some places, like Colorado, new laws are even being considered that make it legally and administratively easier for artists to run their businesses as fundamentally collaborative enterprises.

The point is that working together in this way can produce some of the same feed-mediating and algorithmic effects that artists are chasing through trend simulation campaigns. The difference is that this kind of momentum is rooted in real relationships, mutual support, and some baseline commitment to a shared ecosystem. Just like democracy, working together is hard work. In the music context, it may never fully match the scale and speed promised by manipulative marketing, and the house usually still wins. But for artists and labels interested in a different logic of being that is more about building durable circuits of reciprocity, trust, and shared investment, this is a path worth considering.

Some Music!

A total drivin’ round, windows down, spring is in the air bop. Jane Inc. rules!

Last week I posted a note here asking whether anyone had written an analysis of what these expected inflation shocks might mean for music.

I got a few good responses from some smart folks.

made several points that feel broadly correct. First, wars and periods of uncertainty don’t necessarily hurt music consumption overall. If anything, people often lean more heavily on music during crisis. However, they may also do so in a very particular way, by turning toward familiar and nostalgic music. When times are tough, folks are more likely to listen to ‘A Change Is Gonna Come’ to feel better, than they are to fire up the New Music Friday playlist and seek out challenging, recently released music.

Patrick also pointed out that vinyl and physical segments in music look particularly exposed, with commodity and manufacturing costs rising.

Live touring could face pressure too, especially in markets where promoters and festivals rely on imported headline acts. wrote an excellent piece focused specifically on live, laying out the full logic of how inflationary pressures will ripple across the industry via fuel, transport, food, and basic touring inputs.

What I want to build on, though, is the broader distributional picture. Ben already points to some of it with respect to live, where independent artists and certain regional markets (Europe and Asia) may get hit hardest, but I think the same basic pattern could extend across the wider recorded music economy. Even if “music” in the aggregate proves resilient, that resilience may rest on outcomes that are highly unequal.

Crisis favours catalog

On streaming, the likely beneficiaries of uncertainty are bigger or mid-sized established artists with strong catalog. If precariousness pushes listeners toward familiar, nostalgic music, then the long-tail, power-law distribution that already generates inequality in music (and particularly wrt digital platforms) will likely intensify, with more attention and revenue flowing toward established catalog. That mostly benefits majors, private equity, and the subset of larger or more consolidated indies with meaningful catalog depth, while reinforcing a broader trend in which investment in older music comes at the expense of the future sustainability of music itself.

The impact on streaming may also have a geographic dimension. Western subscription markets are already nearing saturation, so a lot of future streaming growth has to come from elsewhere. But many of the regions that matter most for that next phase of growth, including parts of Asia, MENA, and Africa, could be hit especially hard by energy and inflation shocks linked to the war in the Middle East vis a vis geographic proximity + related oil dependence + broader macro spillovers. That could dampen consumer spending and slow subscription growth in these regions, dragging on the pace of global streaming growth overall.2 At the same time, as Patrick pointed out, music’s role as a source of comfort during conflict may provide some counterweight to these downward pressures and temper some of these effects.

Inflation and the super-fan economy

Aside from streaming, what interests me most in relation to inflationary pressures is the “expanded rights + physical” side of the business, which includes merch, brand partnerships, live3, vinyl, and related revenue streams. These are parts of the business most closely tied to the industry’s emerging super-fan approach to driving revenue growth.

MIDiA’s 2025 music industry macro recap noted that expanded rights grew 21.5%, and that for the first time in the streaming era, streaming revenues grew more slowly than the total market because expanded rights and physical were taking on so much of the heavy lifting. Goldman Sachs makes a similar point in its latest Music in the Air report, arguing that the next phase of industry growth will come from a mix of new streaming models, video, and, importantly, via monetizing super-fans. A lot of future growth optimism in recorded music is tied to these areas. This is music’s growth frontier.

But expanded + physical are also among the most vulnerable areas to inflation shocks, since they depend more heavily on discretionary consumer spending. Even if music fans keep their streaming subscriptions through tough times, they may cut back on vinyl box sets, merch, premium fan memberships, and other ‘extras’ in the budget. At the same time, those same parts of the business are more exposed to rising commodity, manufacturing, shipping, and fuel costs.

Again, the distributional point is key. Expanded rights, physical, and direct-to-fan channels (like Bandcamp) are especially important for smaller artists and labels, precisely because streaming economics are already so poor for them. If those areas weaken too, then another of the few places where smaller artists still had some room to build meaningful income starts to narrow.

Then we add live on top. Another inflation shock could be the proverbial straw that breaks the camel’s back, pushing touring further out of reach for small and mid-level artists and cutting off another of the few remaining ways they can still eke out a living.

Altogether, what this points to is a deepening of the the inequalities already built into the music industry. Large rights-holders and firms with capital to absorb higher costs or pass them on may prove relatively resilient, and may even strengthen their position (vis a vis streaming, at least) as listeners retreat toward familiarity. But the part of the ecosystem that allows newer, smaller, and mid-sized artists to develop looks much more exposed. While ‘music’ in the aggregate may hold up, the conditions underlying the viability of the next generation of artists will continue to deteriorate, producing a music economy that becomes even less capable of supporting the artists who are supposed to make its future.

Some music!

It’s been a hard time for the future lately. So much is going wrong in the world, and the trajectory often feels relentlessly bleak.

Sitting with this one gives me some hope. The future is what we build together, and that means we can still shape it for the better.

I was thinking about the individual binds and incentives placed on artists because of a tweet and related article that has been making the rounds:

The article describes the practice of clipping in music marketing, where songs are inserted into short-form videos posted by paid networks of meme, fan, or influencer-style accounts, to simulate organic buzz and drive attention to an artist on social platforms like TikTok.

The first thing that struck me about clipping is that it is, on its face, deceptive. At best, it’s grey-hat marketing, and pretty clearly a shifty way to game the system, and consequently, to undermine trust between artists and fans. To me, this is part of a broader enshittification of the artist-fan relationship, by taking what ought to be an affective connection grounded in art, culture, and shared meaning, and collapsing it into a fully commodified transaction aimed at maximizing financial gain.

Still, whatever one thinks of the ethics, clipping appears to be widely used. The replies to the tweet make that clear:

“Clipping is so big in streaming and hip-hop. Shocked every industry isn’t leaning in more”

“Clipping isn’t new but its adoption in music marketing is worth paying attention to. For artists, the upside is real. Many dislike making content, and clipping reduces that burden. And compared to traditional marketing spend, the costs are lower, which means faster recoupment.”

“Is there a way to do this for writing?”

So, why do artists and music companies use it despite pretty obvious ethical defects? Is this an individual lapse in morality? Or the result of structural forces? Hint: It is almost always the latter w/ capitalism!

Platforms and organizations exploit the structural logic of the attention economy. Because attention operates as a zero-sum field, in order to gain a foothold, one naturally has to leap over others. With this structural incentive built in, it’s a simple hop skip and jump to building systems designed to exploit it for profit.

A key problem is that these zero-sum incentives are fundamentally short-term in orientation. They’re geared toward maximizing immediate individual advantage and capturing ever more attention-market share which, in practice, usually also means greater economic market share. Music manager Sam Alavi captures the bind that artists face well: “I don’t know anybody not utilizing [clipping] who’s actually competitive in the marketplace.”

As I’ve written before with respect to catalog acquisitions, a fixation on short-term financial incentives can produce an ouroboros effect, where we consume our own tail, undermining the long-term sustainability of music in exchange for immediate payouts. The same logic applies to clipping.

At the institutional level of major labels and private equity, that logic is structurally embedded and often knowingly pursued by the people in charge. But with practices like clipping, which operate at the level of individual artists and their teams, the tradeoff is more obscured. In that sense, artists may be unwittingly participating in the deterioration of the very ecosystem they depend on.

With a bit of analysis, we can see how the internal logic of clipping undermines the future sustainability of music. The attention economy encompasses the full universe of things people might do with their time, whether watching sports, endlessly scrolling social feeds, gaming, or listening to music. Music is already structurally squeezed within that economy, and is often reduced to an underpaid supplement that helps keep attention locked onto other mediums, especially short-form video.

Clipping deepens that problem. By paying fake influencers to saturate social feeds with inauthentic content, artists aren’t just chasing attention away from one another; they’re also cutting into the limited time and cognitive space audiences could be using to engage with music more directly. In doing so, they help intensify the kind of low-trust environment often described by ‘dead internet theory’, where feeds are dominated less by genuine cultural exchange than by manufactured signals, fake participation, and algorithmically optimized slop.2

This also creates a classic collective action problem that manifests as a race to the bottom. As more artists adopt clipping to remain competitive, the wider attention environment becomes more degraded, while any advantage gained from the tactic is also diminished by its general uptake. The result is a self-reinforcing loop in which artists compete for advantage on terms that, collectively, erode the foundation that makes meaningful musical engagement possible.

Clipping is just one of many examples of how race-to-the-bottom dynamics are weaponized within the systems artists are forced to navigate, and by the organizations powerful enough to shape those systems.

Once we know what to look for, we can see the same logic at work across the music ecosystem. Spotify’s Discovery Mode is another example. The program allows artists and rights holders to trade reduced streaming royalty rates for beneficial placement within Spotify’s personal playlist algorithms. It’s a way for artists to buy their way to the front of the attention-line. The structural gains and profits from that tradeoff, however, flow primarily to Spotify and its shareholders. For artists, the deal is much less compelling. If enough artists opt in, the relative advantage of preferential placement is effectively neutralized, since everyone is once again competing on the same terms. When everyone’s super, no one will be. Except now Spotify has succeeded in lowering its overall payment obligations to artists at scale. As with clipping, the apparent short-term individual incentive obscures a deeper long-term tradeoff where participation may offer marginal career gains in the present, while helping normalize a system that erodes music’s long-term sustainability.

With race-to-the-bottom logics seemingly everywhere, the problem is that coordination against them is difficult to organize. The only way out is collective action, but the challenge of coordinating to build the shared values, trust, and long-term strategic horizons necessary for that action is hard. Artists are atomized within their own careers, and pushed to survive as individual competitors inside systems they don’t control. At the same time, austerity and broader economic trends continue to hollow out the infrastructure that makes collective constellations possible. Grassroots venues remain precarious, community institutions are under strain, and arts funding is often among the first things to be cut. The result is a music ecosystem in which the need for coordination is obvious, but the conditions for achieving it are structurally undermined.

Nonetheless, we persist! There are efforts to build this kind of coordination among artists. UMAW, for example, is explicitly organized around a values-based vision of a more just music industry, and has mobilized artists and music workers together around issues including streaming pay, merch cuts, education, and international solidarity.3 The Reset Network is doing similar, values-oriented work in Europe. In policy, the Protect Working Musicians Act in the US points in a related direction. It is an attempt to give independent music creators greater ability to negotiate collectively with dominant streaming platforms and AI companies, rather than leaving them isolated in one-on-one encounters with concentrated corporate power.

Ultimately, coordination and collective action are hard, especially at scale. Organizing is slow, demanding work. But if our analysis of these structural tendencies is correct ( and I believe it is), then there is no real alternative to it. If these dynamics systematically push music toward race-to-the-bottom outcomes, then individual moral choices or isolated acts of resistance will never be enough. The only adequate response is to build the forms of solidarity, organization, and collective power capable of contesting them. At the risk of sounding like a broken record, if we care about the future of music, we have to organize!

Some Music!

I was excited to see that Cole Pulice is coming through Ottawa in May and performing at Fono, a very cool new small venue, community space, and hi-fi listening room opening literally 40 feet from my apartment. My friend and ambient wizard Nick Schofield is opening. I am a lucky guy!

I’ve been obsessed with Pulice’s latest album ‘Land’s End Eternal’ since it came out last year. I love its sense of space, and the textures, and, sacrilege though it may be to tube amp purists, I absolutely adore the digital-artifact sounding direct-in guitar tone that anchors the album.

Late last week, Billboard released a new episode of its On the Record podcast featuring billionaire Mark Cuban discussing music as a capital investment opportunity. The interview offers a glimpse into how private capital, and especially investors from outside the music industry, understand music. The picture it paints is fairly bleak. Here’s what Mark had to say about music:

“Yeah, I think it’s the worst industry ever,” Cuban says with a laugh. “No, it’s probably tied with clothing, branded clothing…that’s the worst, and music is right behind it.” Later, the investor clarified that he was specifically referring to new music and the record labels that release it. To Cuban, older music, also known as catalog, is in a different category entirely — as is music tech.

New art and music and record labels supporting it? Big thumbs down from Mark.

This statement shouldn’t surprise anyone who has been paying attention to investment trends in music, or to the openly anti-art rhetoric coming from many music AI founders and investors. Even so, I think it’s important that we keep calling attention to this language, both for what it reveals about the future being imagined for music, and for what it clarifies about where our efforts need to be directed in resisting this flattening of music and art (i.e. towards producing our own countervailing power).

Subscribe now

What always stands out to me about folks like Cuban is how their language is detached from any real appreciation for music and art. Over and over they frame music instrumentally, as a vehicle for extracting profit, but rarely as a source of beauty, shared meaning, or collective human experience. I mean, Cuban’s own stated favourite song is ‘Life’s Been Good’ by Joe Walsh, a song which is meant to satirize the excess and vanity of a wealthy life, but which he seems to hear instead as a straightforward reflection of his own good life. The song becomes another mirror for his own wealth, which seems to be the only register in which he knows how to hear it.

Some music!

I’ve been digging the new James Brandon Lewis x The Messthetics album. What I love about it isn’t just that it sounds pissed off, but that it meets the current moment with real force. When I shared it in ye old music group chat, responded that it didn’t just sound angry, but also liberating, and I love that. Music like this isn’t for lean-back listening and lamenting the state of things (won’t someone else do something about the world!). It is music that pushes us outward, toward action, organizing, resistance, and seeing our obligations to the collective whole. See you out there ✊ LFG!

Yesterday, I ended by suggesting that artists, workers, and independent music companies need to develop a “coherent politics of shared values and mutual accountability” that is capable of undergirding a coalition with the power to shape the music ecosystem. Essentially, I argue for the necessity of class formation within music as a precondition for producing artist and worker centric reforms.

Another useful way to frame that argument is through the ever-present debates within music over what “independent” is supposed to mean. There is a substantial body of writing on independent music that traces its history, documents its values-based ethics, and contests the meaning of the term itself. I’d recommend reading everything by the very excellent music journalist Simon Reynolds as a starter pack on the history (at least in the UK). There are also more recent essays that clearly articulate the values historically associated with independence, and demonstrate how some music companies that identify as independent have sought to operationalize those values in practice.

Long story short, I’m not the first person to suggest that ‘independent’ in music is a term grounded in a specific history and values, or that that the term has gradually been expanded in ways that detach it from that context.

The problem is straightforward to diagnose. Defining independent music as a negation, meaning “not major”, has the effect of broadening the term to the point of incoherence. The range of business models, politics, and institutional commitments contained within that definition is just too broad for the term to retain any genuinely cohesive meaning. The issue with this is that it leaves the field open for any non-major music company to co-opt the term as a form of brand marketing. In this way, the values-based history of independent music, absurdly, becomes weaponized as a marketing tool available to almost any music company on the planet, regardless of whether they are aligned with, care about, or operate according to those historical values.

This definitional looseness also muddies analysis. Take Concord’s acquisition of Ninja Tune last week. How are we supposed to read a deal like that amid broader concerns about consolidation in music? Consolidation within the “independent” sector carries many of the same risks as consolidation by the majors. But unless we have a clearer and more systematic understanding of the values and practices that actually structure these firms, the label “independent” tells us very little. On its own, it doesn’t provide analytical leverage.1

Altogether, this creates a predictable opening for free-riding. The values-based history of independent music is converted into a kind of transferable moral and commercial resource, available even to firms that neither share nor enact those values. We might call this a “normative halo” or, as my partner suggested more alliteratively, a “values veil”, comprising a layer of market-enhancing ethical legitimacy that offers reputational benefit while requiring very little in return.

This is the problem field. “Independent” once carried a more specific ethical and political meaning, but now often functions as a floating signifier, free for use by almost anyone (except the majors).

There is already a good deal of writing that diagnoses this slippage. What remains less developed is the question of how to solve it. How do we recover the collectivist and humanist values historically associated with independent music while also preventing those values from being reduced to empty marketing?

The simplest response is to make free-riding harder. If the term “independent” has been hollowed out, then one way to reclaim it is to draw firmer boundaries around it. This means more clearly codifying the values it represents, and developing mechanisms through which companies claiming the term can be held accountable to those values. If “independent” is going to function as more than a marketing asset, it has to impose obligations as well as confer benefits.2

That, in turn, brings us back to the argument from yesterday. What is required is not just better language, but the willingness and resources for a coalition of stakeholders (artists, songwriters, indies, other?!) to do the hard work of articulating a shared values framework, defining what this means in practice, and building the political strategy needed to advance an agenda grounded in it.

In other words, the task is not only to describe independence differently, but to organize it. We can continue to debate what “independent” means, but until that meaning is clearly defined and backed by structures of accountability, the term will remain unstable and open to appropriation by anyone who wants its benefits without accepting its obligations.

Some music!

Shane Parish is a guitar god. If you haven’t you should dig in to his full catalog. He’s also a touring member of the Bill Orcutt Guitar Quartet, who are total shredders🎸.

His recently released album, ‘Autechre Guitar’ sees him re-imagining Autechre songs for solo guitar 🤯. It’s wild stuff. Parish’s own description of his version of ‘Slip’ gets at the slipperiness perfectly: This main riff is a 29 beat cycle over a 4/4 beat. So the beat “slips” when the accent shifts and the melody seems like it never repeats.

This quality output is partially how I found myself watching and listening to this exceptional live recording, out now on IA, from the composer and horn-player Rob Mazurek and his trio (w/ Mikel Patrick Avery & Matthew Lux). If you have 36:08, you should really sit down and give it a front-to-back w/ good headphones or stereo:

Aside from the musical performance itself, though, something else popped out to me when I was browsing the video description:

Catch that? Buried at the bottom:

In other words, if you want to stream this album interactively, Qobuz is the only place to do it. If you’re a Spotify, Apple Music, Tidal, etc. user, you won’t find it in your app.

This stood out to me not simply as a distribution detail, but as a small example of how relationships in music might be organized around shared values.2

When I dug in, I realized the release is part of an ongoing series of co-presented live shows between IA and Qobuz, along with a related run of exclusive live albums. The partnership makes sense. There is an obvious aesthetic alignment around high-quality audio, which is especially lovely for live recordings like the IA x Qobuz series.

But beyond aesthetics, what seems worth highlighting is that IA and Qobuz also appear to share some artist-friendly values and commitments. Qobuz has structured its platform in ways that are more artist-friendly than many of its competitors. It has no free, ad-supported tier3, functions as both a streaming service and a marketplace for direct purchases, emphasizes human curation and editorial, and reports generating “on average five times more revenue per user than the market average.”4

Toward a Politics of Independent Music

None of this is to suggest that International Anthem and Qobuz are alone in pursuing this kind of values-aligned collaboration. The larger point to take away is that if we want music to be organized differently, we need to be more explicit about the values we want it organized around, and more willing to act on them.

In other words, we need a politics of (independent) music that includes a clear sense of principles, commitments, and forms of accountability that should guide decision-making in the first place.

International Anthem is a particularly useful example here because it appears to understand itself as more than just a market actor. It presents itself as part of a broader cultural and social project, which makes it a helpful case for thinking about what it would mean for practical business decisions to be guided by an articulated politics rather than by convenience alone.

International Anthem announces its values and politics right up front on the ‘About’ page of its website:

Notably, its mission statement doesn’t stop at aesthetic concerns alone, but articulates a distinct politics of collectivism and sustainability as a foundation for boundary-defining music. They also don’t limit their scope simply to artists, but extend outward to a community of workers, expressing solidarity openly.

Among my Board colleagues at experimental music presenter Debaser, I’m known as the guy who is always droning on about how an organization’s mission and vision should guide practical, values-based decision-making and, on the other end, enable accountability to its community. I really believe this. By publishing its mission statement publicly, International Anthem is in turn creating a public mandate for it’s work, and with that, the conditions for public accountability.

From Shared Values to Collective Power

Recently, I’ve been thinking and writing about the need for some form of collectivization or coalition-building among values-aligned actors in the music industry, with the idea being to build clear, durable power across artists, workers, and independent music companies. However, a coalition like this needs to be grounded in a coherent politics of shared values and mutual accountability.

Leveraging this kind of power would require coalition partners to make tradeoffs at different times. For that to work, participants in the coalition need to see one another not as competing interests with occasional overlap (i.e. one-off policy-issue dance partners🕺), but as part of a common collective with a shared set of stakes. We might even call this a kind of “class” formation within music, with a co-developed, material, values-based politics from which to act and organize.

What does that actually look like? What are the specific values embodied by such a coalition, and how can they be articulated in ways that are meaningful rather than vague branding? That is still an open question for me, and one I think the music sector needs to engage much more directly.

And of course, I’m idealizing here. A coalition of this scale will require organizing, trust-building, strategic planning, and forms of coordination that can turn shared values into long-term power.

Still, the point is not to arrive immediately at a finished doctrine or to impose a fixed set of values from above. It is to be more honest about the fact that the debates we’re already having about music are, at core, debates about values. They are about how the music ecosystem ‘ought’ to be and be built together (the ‘normative’), even if we’re talking about practical things like partnerships, who buys which companies, and what that means in terms of outcomes for artists and the music ecosystem.

Seen in that light, the significance of public mission statements like International Anthem’s, or of values-aligned partnerships like its collaboration with Qobuz is that they make a different set of values and commitments legible. And if independent music is going to build durable collective power, that kind of political, moral, and practical legibility is where the work has to begin.

Some music!

You’re probably already listening to Rob Mazurek and co., but in case you need some additional music (and who doesn’t?!), SML’s ‘How You Been’, which is also out on International Anthem, was my favourite album of last year.

Born of improvised sessions among the group, then meticulously chopped + rearranged through deep collaboration, this is music made from trust & shared risk. Is this socialist music? Maybe! Either way it rips.

As you can see, the majors have been on a buying spree in recent times. Alongside the map, Ben convincingly argues that a substantial share of this investment is concentrated in distribution and related infrastructure, reflecting a major label strategy to secure greater leverage over DSPs and exert more control over the “highway” through which independent music reaches audiences. He further suggests that this positioning matters even more in an AI-driven future, where control over catalog access, attribution systems, and payment rails may become increasingly strategic. As Ben writes, “It’s a worrying trend that leaves fewer and fewer independent paths to releasing music.”

What Ben is documenting is a form of horizontal integration, where each major is expanding its ownership of complementary businesses operating at roughly the same point in the value chain. All three majors now hold multiple independent music distribution-related businesses, providing control over an increasingly important layer of the pipeline through which independent music is delivered to platforms, monetized, and brought to market. That, in turn, gives them not only another source of revenue, but also greater strategic leverage over a part of the ecosystem that more and more artists now rely on.

When I saw Ben’s graphic, it struck me that it could be useful to add another layer of categorization by locating each acquisition and investment within the broader music value chain:

To do that, I divided the music value chain into eight constituent categories and then classified each deal highlighted by Ben accordingly23: (1) Creation & A&R; (2) Production Tools; (3) Rights Management / Attribution; (4) Catalog Investment / Rights Finance; (5) Artist Services & Marketing; (6) Labels / Label Services / Distribution; (7) Fan / Community / Commerce; and (8) Consumption / Listening.

Even a quick glance at the map shows that, while there are clearly some parts of the value chain attracting heavier investment (like distribution, as Ben notes), the majors’ acquisitions and stakes are spread across nearly the entire chain, from creation to consumption.

This is vertical integration, where companies attempt to own or control multiple stages of the supply and value chain. This is not new to those of us who have been paying attention to the music industry recently, as indies assembled a broad coalition to oppose Virgin Music’s acquisition of Downtown on competition grounds, including vertical integration concerns.

While there is nothing especially novel about pointing to vertical integration as a problem in music, investment maps like this still serve a useful purpose. They make visible how the major music companies are deploying capital through investment to consolidate control across the full value chain, and with it their power to shape the terms on which artists, workers, and smaller firms operate.

These are not speculative risks; they are well-recognized competition concerns across the United States, the European Union, and Canada.4 We know that vertical integration can, for example:

• Produce foreclosure, cutting rivals off from key inputs, services, datasets, or routes to market (e.g., an independent artist or label losing access to a key distribution or playlisting channel.)

• Enable self-preferencing, allowing integrated firms to favour their own repertoire and affiliated services over independents (e.g., a platform giving better placement to music tied to its own label group, or in which a label group has equity stakes.).

• Increase bargaining leverage by bundling multiple services across the chain into offers that artists and smaller firms may have reduced ability to refuse (e.g., offering distribution, marketing, and advances as one package.).

• Concentrate data and attribution power, giving a handful of firms greater control over the metadata, listener information, and recommendation systems that increasingly determine visibility and payment (e.g., controlling the financial, accounting, metadata and recommendation systems that determine visibility and royalties).

• Facilitate cross-subsidization, allowing dominant firms to use profits from one line of business to underprice competitors in another (e.g., using profits from a major catalog to underprice an affiliated marketing or label-services arm.).

Although the European Commission ultimately approved the Downtown transaction, the campaign was not for nought. As a condition of approval, Downtown was required to divest Curve Royalty Systems, a platform central to the financial operations of many independent music companies. That is why, fortunately, Curve does not appear in this map.

This demonstrates that even where regulators approve a transaction, coordinated opposition can still shape the outcome, narrow the scope of a deal, and clarify the stakes for the wider sector. In this case, the requirement that Downtown divest Curve was not a trivial concession. It reduced one of the clearest risks of the integration by helping prevent sensitive financial data from independent music companies using Curve from flowing into major-label hands, where it could have been used to deepen their informational advantage and bargaining power.

Ultimately, what investment maps like these make clear is that the issue is not just one of individual acquisitions or isolated markets. There is a broader pattern of consolidation and integration — both horizontal and vertical — across music, through which the majors and other large holders of capital are deepening their power over the infrastructure that increasingly determines how music is made, circulated, and paid for. If that trend is to be challenged, artists, workers, and independent firms will need to strengthen their coalition, while pooling resources for the research, policy and regulatory advocacy, public education, and organizing needed to resist it.

Some Music!

Since we’re talking about how vertical integration can lead to the disintegration of the institutions needed for a sustainable music ecosystem, it seems only fitting to turn to William Basinski’s ‘Disintegration Loops’, a dark, haunting record that feels uncomfortably well suited to the present moment.

For anyone interested, here is the Excel file in which I classified each investment:5

Vertical Integration In The Recorded Music Industry Database

7.45KB ∙ XLSX file

Download

Download

I’ve got a new essay out there today on investment power in music in the age of AI—who controls it, how that power shapes everything downstream, and what artists, workers, and independents can do to build real leverage to influence the future of music.

It’s a response to ’s excellent essay from last week, which is well worth reading. I find it genuinely motivating to be in public conversation about the values and strategies that artists, workers, and indies can collectively harness to push music toward a better future, especially in a moment when the world (and the music industry) can feel particularly bleak.

Head over, give it a READ, subscribe to MUSICx, and if it sparks something for you, jump into the comments.

Some Music!

My essay is a reply to Darius who is also the founder of Jagjaguwar, the dope independent label behind many of your favourite artists (Angel Olsen, Lonnie Holley, Bon Iver, and more).

Jagjag’s latest release is this new jam from Gia Margaret. Her upcoming album marks a return to proper lyrics-based songwriting, after a turn to gorgeous piano-forward ambient compositions.

This new one exudes spring vibes. The weightless groove and bassline. Those little short-burst descending synth arpeggios in the chorus. It all makes me want to go stand out in the above zero weather and stare at the sun.

My recommendation: Take it for a walk in your headphones when the sun is out. You won’t regret it.

In drafting that piece, I ended up cutting a few sections that didn’t quite fit, including the one below on the rapid growth of catalog investment, aka investment in older music.

Then, today I read Warner Music CEO Robert Kyncl’s latest letter to shareholders, where he cites this statistic: '“just 27% of U.S. music streaming consumption in 2024 came from new releases, down from 45% a decade ago.

That statistic made me want to pull the cut section back out, because it gets at a different but related question about investment in music: What happens when more and more music capital is directed toward the past rather than the future?

Here’s that short section that I chopped. It reads a little wonky out of context of the full essay, but that’s life in the blogosphere🤘:

Investing in the past

The most visible symptom is the catalog acquisition boom, a multi-billion-dollar wave of capital that has seen at least $20.4 billion flow into music rights acquisitions since 2019, according to WIPO.

Catalog investment is not entirely new. The modern era of financialized music catalog investment arrived with the dawn of the new millennium, and has steadily grown over the years since.1 While flows briefly dropped off as interest rates rose and money became expensive near the tail end of the global pandemic, catalog-oriented investment has since recovered and is going stronger than ever today. In mid-2025, Warner Music Group launched a $1.2 billion fund with Bain Capital to acquire music rights. Weeks later, Chord Music Partners, which is backed by Searchlight Capital, Universal Music Group, and a constellation of family offices and pension funds, raised over $2 billion in investable capital. Last month, Sony Music Group announced an even larger joint venture with GIC, Singapore’s sovereign wealth fund, targeting $2–3 billion in music rights acquisition. If you’re keeping score, that makes all three major labels in partnerships with multiple billions earmarked to purchase the rights for old music, alongside countless other funds.

The investment logic is clear enough. Catalog royalties provide predictable, securitizable cash flows that are inelastic to movements of the broader market (Even in tough times, listening to Bruce Springsteen slaps, ya know?2). They are the kind of yield-generating asset that financial institutions understand.

Treating music’s back catalog as a financial instrument makes sense if profit is your only criterion and your time horizon is short to medium term.

What this investment doesn’t do is support music’s future. It doesn’t directly fund new music or develop new artists. Even if major labels or streaming platforms choose to reinvest some of those profits in new artist development, they could just as easily return them to shareholders through stock buybacks (like Spotify did recently). Either way, large institutional investors may still extract those gains from the music ecosystem altogether, or recycle them into further catalog acquisitions.

There is a longer-term problem embedded in this logic. Music catalogs generate reliable revenue precisely because they reflect the accumulated, past investments in artist development, including the artists who were developed, the recordings that were made, and the sounds that shaped generations of listeners. But audiences age. The next generation of fans who are forming their tastes now, will want their own artists and music to love. They won’t be served by an industry that mainly invests in music that already exists. The data is already suggestive: streams of current music are falling, while catalog’s share is rising, with at least part of that shift driven by capital flowing into catalogs and the marketing spend that follows, which pushes older music harder and further crowds out the new.

Kyncl quotes

And for good measure, here’s a few more of those bleak quotes from Robert Kyncl’s letter to Warner Music shareholders. It’s like he doesn’t even like music, let alone love it:

“With algorithmic personalization fueling interest in artists from all eras, new listeners of all ages are exploring historic catalogs and embedding clips into short-form videos. Best of all, there’s plenty of room for even more growth, especially as AI begins to enable fans to more deeply interact with the music they love, by reimagining it rather than just listening.”

🥴🥴🥴

“We’re also seeing robust performance from our catalog as we keep brands relevant and growing in the digital era through an always-on marketing philosophy, finding powerful sync placements driving increased engagement for legends like Prince, David Bowie, Fleetwood Mac, and many others. With additional firepower through our joint venture with Bain, we expect to deliver accelerated growth through margin-accretive acquisitions of high quality catalogs, starting in 2026.”

🥴🥴🥴

“Today’s passive listening experiences are rooted in a market-share-based system. In a more lean-forward, interactive world, we are embracing an attribution-based system that will drive value to the most iconic artists and catalogs.”

🥴🥴🥴

So, I’m going to start experimenting with posting short research notes here. The idea isn’t to generate a paid readership. I’m generally in the camp that believes we should be trying to push towards more collective publishing endeavours that can support a sustainable music media ecosystem1 But, the type of writing I want to do here — short, quick, unvarnished thoughts that are looking for feedback and dialogue — isn’t currently a great fit for most collective publications (“publish my half baked ideas” isn’t a good pitch!). And so to get them down somewhere, I will be putting them here.

Forcing myself to write here more often will also (1) force me to undertake better analysis, and test my own (worst) first thoughts. I’m a person who believes writing is thinking, and if you don’t use it you lose it, and I don’t wanna become a dummy, and this will help with that🤓.

I also want these posts to serve as an entry point for readers to respond and to create a space where we can sharpen our thinking together.

With all that said, here’s a lil’ research note about music AI attribution tools!

I’ve been closely following the emergence of music AI attribution tools, including Sony’s recent announcement of its new music AI attribution technology. Briefly, AI attribution systems analyze AI-generated music to estimate which source works — the artists, styles, or musical features — that AI output draws from. They are used to inform decisions about credit, compensation, licensing, rights management, and transparency in AI music systems. In practical terms, when a generative music AI platform outputs a song, an attribution system is designed to assess how much that output may reflect the training data the model learned from, and from whom (is it more Springsteen? or Petty? or your local dad rock band?), and then trigger payments to those artists who influenced the AI generated song (in a technical sense).

I keep coming back to what feels like the most important question for evaluating these systems:

How do they actually work?

A recent Forbes article describing Sony’s new tool says: “For example, the Sony technology can state which musical artists the output resembles and at what percent influence.”

But what does that actually mean? The design space for “attribution” is huge. What counts as resemblance? How is “influence” being measured? At what level — melody, timbre, style, structure, lyrics, production choices? How we measure these things makes all the difference in who is ultimately gonna get paid and how much.

I keep returning to this excellent Water & Music piece (shouts to the homies, Cherie Hu, Yung Spielburg and Alex Flores for doing the lords work!) which maps out some of the major approaches to designing attribution systems in music AI.

My main takeaway from learning about these systems isn’t ‘Is attribution possible?’ Obviously it is, and can be done in many different ways.

Once we accept this, then we need to reframe the whole thing as a political question: “What is attribution?” in the context of designing systems where attribution is the metric used to decide compensation. The question is political, not technical.

I think of it a bit like electoral systems. There is no single “best” system. There are different systems that produce different outcomes because they embody different values. The question is not purely technical, but rather is about governance, power, and whose interests are being prioritized.2

Music AI attribution systems are no different. Who gets paid is a function of system design, and different designs will distribute value differently.

My worry, of course, is that if technologists (or wealthy folks with capital) are the ones deciding which attribution systems become the standard, rather than artists and music workers, we are likely to end up with some wonky, reverse-robin hood style attribution systems, that funnel money to the largest artists and music companies, while entirely undercutting the value and livelihoods of most artists and music workers.

Some Music!

If you know anything about me, it’s that I love music. I love to share it and talk about it, and listen to it. Art rules! Music rules! It makes life better! I even make a point to invert the commodified logic of LinkedIn networking, and frequently just post good music to my feed there.

So, I’ma probably end these research notes with a link to some music I’ve been loving lately.

Here’s one of my fav pieces of music that I come back to again and again. Jim O’Rouke is one of the GOAT’s. Here, he presents a a 30-minute long, live cover of Tracy Chapman’s Fast Car. It requires some real patient listening, and gets pretty drone-y in the middle section, but emerges into a beautiful butterfly that is worth the wait!

PS. I’m also part of a group, alongside and some others, that have a Signal chat where we simply share music we’re listening to with a short sentence about why it rules. The group is small currently, and tastes are diverse. If you want in, hit me up in the DMs, and I will send over a link 🤘.