MBW Views is a series of op-eds from eminent music industry people… with something to say. The following MBW op-ed comes from Andreea Gleeson, former CEO of TuneCore and founder of AGA (Andreea Gleeson Advisory), where she advises companies and investors on growth, innovation, and strategic transformation across music, media, technology, and the creator economy.
Here, Gleeson argues that AI’s biggest challenge isn’t the technology itself, but building the interoperable commercial infrastructure needed to ensure creators are properly credited, compensated, and protected.
Over the past few months, the music industry has quietly crossed an important threshold. Not another lawsuit. Not another congressional hearing. Not another debate about whether AI is a threat or an opportunity for music. Instead, we’ve entered the commercialization phase of AI.
In just the last several weeks, Deezer announced that more than 90,000 AI-generated tracks are now uploaded every day, representing over half of all new uploads to its platform. TIDAL announced it will automatically identify fully AI generated recordings and exclude them from royalty bearing streams. The global record industry has proposed standardized AI labels for streaming platforms.
IFPI has also rolled out chart eligibility principles across its global network of official music charts, using those AI labels to determine how AI Assisted and AI Generated recordings will qualify for chart inclusion. Spotify and Universal Music Group along with Merlin unveiled licensed AI tools that allow fans to create authorized covers and remixes from participating artists. Meanwhile, Congress continues advancing the bipartisan No Fakes Act to protect voice and likeness.
Viewed individually, these announcements may seem unrelated. Together, they point to something much bigger: the commercial rules for AI music are already being established. If history is any guide, those rules will shape the industry long before legislation catches up.
We’ve seen this movie before
Spotify fundamentally changed music consumption 10 years before the Music Modernization Act became law in 2018. YouTube‘s launch of Content ID in 2007 similarly transformed one of the industry’s biggest copyright challenges into one of its most important attribution and monetization systems, years before policymakers fully understood the implications of user generated content. In both cases, the marketplace standardized commercial models first. Regulation followed.
AI appears to be following the same path.
The next chapter won’t be defined by the technology itself. It will be defined by the infrastructure we build around it and whether that infrastructure properly credits, protects, and pays creators.
My perspective comes from spending more than a decade at one of the music industry’s key commercial crossroads: distribution. When I was CEO of TuneCore, I worked closely with DSPs, labels, creator tools and technology partners to help independent artists bring their music to market. Sitting at the distribution layer gave me a unique vantage point into how missing information upstream often translated into missed attribution, lost monetization and fewer opportunities downstream. Over the past three years, I also worked alongside many of these same companies as they began experimenting with AI, providing a front row seat to how today’s commercial frameworks have started to take shape.
I experienced that shift firsthand in 2023, when generative AI first burst into public consciousness and uncertainty dominated nearly every conversation. At the time, I partnered with Grimes through TuneCore to launch one of the first frameworks for responsibly distributing AI collaborations. Her proposal was remarkably simple: creators could use her AI voice model, but only with her permission and only if revenues were shared. Rather than rejecting AI, the framework established principles that continue to underpin many of today’s commercial discussions: consent, control, compensation and transparency.
Looking back, what strikes me most isn’t that the framework answered every question. It didn’t. It’s that the industry didn’t wait for legislation before beginning to experiment. Artists were already exploring AI as a creative tool. Technology companies were building products. Distributors were developing policies. Platforms were adapting their business models. The market started solving problems while lawmakers were still defining them, and many of today’s most significant developments continue to reflect those same underlying principles.
Consumers Are Sending a Similarly Nuanced Message Too
Earlier this year, Luminate‘s Generative AI in Music report found that overall interest in AI generated music declined from a net negative 13 percent in May 2025 to negative 20 percent by the end of the year, with the sharpest decline among Gen Z and Gen Alpha. But Luminate’s latest 2026 Midyear Report and their AI & Media: Audience Attitudes deep dive suggest the conversation is becoming more sophisticated. Rather than rejecting AI outright, consumers are differentiating between AI that enhances human creativity and AI that replaces it. One in three U.S. music listeners say they are comfortable with AI creating song instrumentals, while 46 percent are uncomfortable with AI creating an entirely new song performed by an AI voice.
The same report found that creators are embracing these tools more readily than the general public. 54 percent of U.S. musicians report positive feelings toward AI music tools, compared with 35 percent of non musicians, and 18 percent already use AI to edit or remix existing music. Together, the findings suggest consumers and creators aren’t rejecting AI. They are asking for transparency, authenticity and human agency while embracing new ways of creating, collaborating and participating with music.
These two signals, the industry’s rapid experimentation and consumers’ growing demand for transparency, are beginning to converge. Together, they point toward the next challenge: how the commercial infrastructure around it should be built. The answer begins by recognizing that AI is not a single technology or a single market. It is an interconnected value chain.
The Four Interconnected Layers of the AI Music Value Chain
If AI is entering its commercial era, it’s important to recognize that no single company or technology will define it. Instead, commercialization is taking shape across four interconnected layers of the AI music value chain. The first is creation, where music is made. The second is distribution, where music enters the commercial marketplace. The third is attribution, where provenance, transparency and ownership are established. The fourth is consumption, where streaming platforms determine discovery, labeling and ultimately monetization.
Each layer depends on the one before it. Information captured during creation ultimately influences everything downstream, from attribution and royalty payments to consumer trust. Looking at AI through this lens shifts the conversation away from individual products and toward the infrastructure that will ultimately determine how value flows through the music ecosystem.
One of the biggest misconceptions about AI is that it primarily encourages replacement. Increasingly, we’re seeing the opposite. It encourages participation. Several years ago, MIDiA Research predicted that music would evolve from static music to dynamic music, where fans don’t simply consume songs but actively interact with them. That future is already beginning to emerge.
One of my favorite examples comes from legendary Chicago house vocalist Robert Owens. Rather than treating AI as something to fear, Owens partnered with Voice-Swap, Beatport and LabelRadar to invite producers and creators around the world to create music using his licensed AI voice model. The goal wasn’t to impersonate him. It was to collaborate with him. The winning entries weren’t celebrated because they fooled listeners into believing Robert Owens recorded them, but because they expanded what collaboration between artists and fans could look like.
Spotify’s recently announced partnerships with Merlin and Universal Music Group to enable licensed AI covers and remixes from participating artists and songwriters takes this idea one step further. It productizes the very behavior MIDiA envisioned years ago of a bifurcated market, creating a commercial framework where fan participation can be licensed, monetized and shared. In many ways, it echoes YouTube’s introduction of Content ID. What initially appeared disruptive ultimately became one of the industry’s largest monetization systems because the right commercial infrastructure was built around it. AI has the potential to follow a similar path, not by replacing artists, but by creating new ways for artists and fans to create value together.
That brings us back to the first layer of the value chain: creation.
Much of today’s conversation focuses on AI detection, and for good reason. Detection technologies are becoming increasingly sophisticated, helping distributors and streaming platforms identify AI generated recordings, protect rights holders and improve trust across the ecosystem. But detection has an inherent limitation: it begins after the music has already been created.
Imagine serving a complex meal to a chef and asking them to identify every ingredient and every step used to prepare it. An experienced chef might come remarkably close, but they are still reconstructing the recipe after the fact. Now imagine the sous chef writing down the recipe as the meal is being prepared. Every ingredient, every measurement and every substitution is documented as it happens. That is the difference between detection and provenance.
“This is where digital audio workstations, or DAWs, become one of the most overlooked pieces of the AI ecosystem. AI is no longer confined to standalone applications. It is increasingly embedded directly into professional creative workflows through vocal modeling, stem separation, mastering, songwriting assistance and other production tools.”
This is where digital audio workstations, or DAWs, become one of the most overlooked pieces of the AI ecosystem. AI is no longer confined to standalone applications. It is increasingly embedded directly into professional creative workflows through vocal modeling, stem separation, mastering, songwriting assistance and other production tools. As AI becomes part of the creative process, DAWs become the ideal place to capture trusted provenance. Rather than asking downstream detection systems to estimate whether AI was used, the software itself could securely document which AI tools were used, what was human performed and even verify when no AI was used at all.
Think of it like the organic sticker on a banana. Consumers don’t inspect the fruit and guess whether it’s organic. They trust a certification system that followed the product from its origin. Music may ultimately require something similar, not because listeners need technical metadata, but because trust increasingly depends on verifiable provenance.
Metadata Becomes Money
That distinction is becoming increasingly important as AI labeling begins influencing commercial outcomes. TIDAL’s recent decision to identify fully AI generated recordings and exclude them from royalty bearing streams demonstrates that AI classification is no longer simply informational; it is becoming economic. At the same time, the industry is exploring standardized AI labels across streaming platforms, an important step toward greater transparency.
IFPI has also introduced chart eligibility principles across its global network of official music charts that use AI labels to determine how AI Assisted and AI Generated recordings qualify for chart inclusion. Chart eligibility extends well beyond industry recognition. It influences visibility, promotional opportunities, consumer discovery, and a range of downstream commercial benefits that often accompany chart success. As AI labels begin informing both monetization and chart eligibility, the accuracy of the underlying metadata becomes increasingly consequential.
But labels alone are only the final output. Without trusted provenance upstream, labels remain declarations rather than verifiable facts. Who provided the information? Was the AI model licensed? Did the artist consent? How much of the recording was AI generated? As AI labeling begins influencing royalties, licensing, recommendation systems and consumer trust, those questions become increasingly important.
Fortunately, many of the building blocks already exist. The MIDI Association’s work around MIDI 2.0 points toward a future where creative tools, distributors, detection technologies and streaming platforms exchange standardized provenance throughout the life of a song. Creation metadata flows into distribution. Distribution informs attribution. Attribution powers labeling. Labeling supports monetization. The value chain becomes connected.
That is why I believe AI is rapidly becoming less of a copyright challenge and more of an interoperability challenge. Detection technologies will remain essential for validating provenance and identifying bad actors, but the strongest ecosystem will combine verified information captured at creation with independent verification downstream. Together, those systems create something the music industry has historically struggled to achieve at scale: confidence. Confidence that artists receive proper credit, rights holders receive proper payment, consumers understand what they are hearing, and AI can expand creativity without eroding trust.
From Innovation to Coordination
The encouraging news is that the industry doesn’t have to start from scratch. Across every layer of the AI music value chain, meaningful work is already underway. Streaming services are developing new approaches to AI transparency and monetization. Labels are negotiating licensing frameworks with AI companies. Distributors are establishing policies around AI submissions. Detection companies continue advancing attribution technologies. Standards organizations like The MIDI Association are building interoperability frameworks, while legislators continue pursuing protections such as the No Fakes Act.
The challenge is no longer innovation. It is coordination.
Each of these initiatives addresses an important piece of the puzzle, but none can establish a connected AI ecosystem on its own. Provenance captured during creation has limited value if it cannot flow seamlessly through distribution, attribution and ultimately monetization. The next phase requires connecting these efforts into shared commercial infrastructure rather than continuing to build them in parallel.
One promising example is the Music Tech Coalition initiative being spearheaded by Peter Brown of Venable. As Peter shared with me recently, “Rather than replacing existing work, the Coalition’s goal is to connect it, bringing together platforms, labels, distributors, technology companies, standards organizations, trade bodies and policymakers to align around practical interoperability. This is not just about the technical standards and agreements that exist, but about industry leaders willing to guide how it all works together and assist implementation across the ecosystem.”
“The next chapter won’t be defined by one AI model, one lawsuit or one piece of legislation. It will be shaped by the systems we build between creators, creative tools, distributors, attribution technologies and platforms.”
Perhaps the clearest sign that the industry is ready for this conversation comes from the recent A3E survey, commissioned by Venable. Nearly 89 percent of respondents believe the industry lacks sufficient coordination around interoperability, transparency and trust. More than 92 percent expressed interest in participating in future coalition discussions, with metadata standards, provenance and interoperability emerging as the highest priorities.
Those findings reinforce what I’ve observed throughout the past several years. The industry isn’t debating whether trust matters. It is increasingly aligned on how to build it.
If the past two decades have taught us anything, it is that the greatest opportunities in music often come from building shared infrastructure. Streaming didn’t succeed because of licensing agreements alone. It succeeded because an entire ecosystem evolved around common commercial frameworks. Content ID wasn’t simply a technology. It became an industry standard for attribution and monetization.
AI now presents the industry with a similar opportunity.
The next chapter won’t be defined by one AI model, one lawsuit or one piece of legislation. It will be shaped by the systems we build between creators, creative tools, distributors, attribution technologies and platforms. The companies that create the greatest long term value may not be those with the most advanced AI, but those that help establish the trusted infrastructure through which AI can scale responsibly.
Congress will eventually write laws. But the commercial foundations of AI music are being built today.
The opportunity before us is to ensure those foundations are interoperable, transparent and artist centric. If we get that right, AI does not have to become a race to the bottom. It can become the catalyst for a more collaborative, economically sustainable music ecosystem, one where artists are properly credited, fairly compensated and empowered to participate in the next era of creativity.
Music Business Worldwide
