I welcome the decision by some music platforms to label fully AI generated music.

There is a real problem. A person can generate thousands of tracks in seconds, upload them in bulk and collect tiny payments that become profitable only because almost no time, labour or thought was invested.

Deezer now receives almost 75,000 fully AI generated tracks every day, representing 44 per cent of all new uploads. It also reports that 85 per cent of the streams generated by such tracks in 2025 were fraudulent. Label them. Remove manipulated streams. Keep automated spam out of recommendations. That makes sense.

But then comes the much more dangerous term: “AI assisted music”.

What does that even mean?

A composer may write every note, every chord and every word, then use AI to change the instrumentation. A synthesizer once allowed us to turn a keyboard into strings, brass or sounds that had never existed before. Today AI can help us move between instruments and voices with even greater freedom.

A composer can upload their own voice and use it to sing what they cannot physically perform. Is that fundamentally different from pitch correction?

One experienced producer has estimated that 99 per cent of modern pop uses corrective Auto Tune. It is an industry estimate, not a scientific measurement, but few producers would dispute that pitch correction has become close to universal. A doctoral study also found that 93 per cent of its respondents used digital audio workstations. Another study of 2,399 Billboard songs found drums in 96 per cent and synthesizers in 29 per cent.

If we start excluding music because technology helped realise it, there will not be much music left.

Perhaps one per cent.

Come to think of it, that is roughly the number of artists who already seem invited to share most of the money. A UK Competition and Markets Authority study found that more than 60 per cent of streams went to music from just 0.4 per cent of artists.

But jokes aside, AI is also part of a democratisation of music creation.

A person who cannot play an orchestra can still compose for one. A creator with a physical disability can turn an idea into sound without first mastering an instrument their body may not allow them to play. Someone without money for musicians, studios and engineers can hear an arrangement, test an instrument or receive advice while remaining the person who wrote the music and the lyrics.

Technology can remove the barrier between imagination and sound. It does not automatically remove the human being.

Of course, the industry must distinguish this from automated music factories. But today’s detection technology cannot do that with certainty. Deezer has acknowledged that detection rates can fall sharply when systems encounter new models or unfamiliar data. Researchers also warn that AI music detectors can produce both false positives and false negatives and should not be used alone for decisions affecting artists or distribution.

We have already seen music penalised for alleged artificial streaming even when artists say they neither ordered nor knew about it. Independent creators are often left with little evidence and limited ability to defend themselves.

Do we really believe the same system will give a red card to the world’s largest artists if AI has been used somewhere in their production?

Probably not. The favourite team, forgive me, the favourite artist, will most likely keep the royalties and collect another medal.

So yes, label fully AI generated mass production.

But do not confuse the tool with the creator.

And do not reduce a human composer’s value because technology helped their music travel the final distance from imagination to sound.

Applause is wonderful.

Recognition matters. Encouragement matters. We all appreciate seeing initiatives that appear to move our industry in the right direction.

But what happens when the applause itself becomes the incentive?

A recent example that generated a lot of applause is TIDAL’s decision to label what it considers AI generated music and exclude it from royalty payments. This feels reasonable. If content is automatically generated at industrial scale without meaningful human creativity, many would agree that it should not simply dilute the value created by human artists.

So far, so good.

The more interesting question is not whether this decision is right today, but what it may enable tomorrow.

History tells us that platforms rarely stop at the first argument. Once a distinction has been accepted, it often becomes the foundation for the next one.

We’ve seen this before.

In television licensing, new competitors were initially welcomed as a way to promote diversity and broaden the market. In Sweden, broadcasters argued that services such as Epidemic Sound increased competition and offered new opportunities for creators. Before long, however, that same development became part of the financial argument in licence negotiations, suggesting that traditional collective licensing should be worth less because some music was now sourced elsewhere.

Internationally, collecting societies such as STIM, PRS and GEMA have similarly faced arguments that parts of platform consumption somehow fall outside the value they represent.

The reasoning evolves.

Today, AI music may simply be excluded from remuneration.

Tomorrow, AI music will almost certainly split into two categories.

One category will consist of vast quantities of automatically generated content that attracts little or no audience.

The other will consist of AI generated works that attract millions of listeners, generate substantial commercial value and become strategically important to platforms.

Those works will inevitably command commercial value and demand commercial recognition. Alongside today’s largest human made catalogues, they are likely to move increasingly towards direct licensing and premium commercial deals.

When that future arrives, we should already be using the technologies that make it fair. Not only to identify AI generated works, but also to ensure transparency, establish provenance and enable fair remuneration for everyone who contributes to the creation of that value.

Meanwhile, the remaining vast pool of creators may once again find themselves sharing a smaller collective pie.

Ironically, those celebrating the first step the loudest may eventually discover that the long term consequences affect them most.

This is not a prediction of doom.

Quite the opposite.

I have never been more optimistic about the technology itself.

We already have the technical capability to move beyond assumptions and towards facts. We can move beyond the outdated pro rata model, where remuneration is largely detached from the actual use of each individual work and instead redistributed from a collective pool that inevitably favours the biggest repertoires. We also have provenance technologies that can establish authorship, timestamp creation and create trustworthy records of creative works from the very beginning.

Solutions such as ProofProfile, developed by New Internet Media, demonstrate that protecting originality no longer has to rely solely on trust. If someone copies your work later, they may imitate it. They can never become the first verified creator.

The same technological shift is also transforming music licensing. As streaming alone becomes increasingly difficult as a sustainable income source for many creators, sync licensing is becoming an essential revenue stream rather than a secondary one.

Here too, technology is opening new possibilities. Platforms such as servo.music are making professional sync licensing accessible beyond the largest catalogues, allowing every song, not only those backed by major players, to compete on its own merits while creators retain control over licensing terms.

The sync market is also likely to be among the most cautious when it comes to AI generated music. Film studios, broadcasters, advertising agencies and brands cannot afford uncertainty around ownership or provenance. As technology increasingly makes it possible to verify where creative works originate and whether copyrighted material has contributed to their creation, transparency will become a competitive advantage. That is one reason why platforms such as servo.music continue to focus on promoting human created music with clear rights and trusted provenance.

Perhaps that is where our applause should be directed.

Not towards intentions.

But towards solutions that genuinely strengthen creators’ positions.

Because when we applaud real results rather than symbolic gestures, innovation accelerates and the entire creative ecosystem benefits.

I’ll applaud that.

#AI #MusicIndustry #Copyright

Maybe that isn’t quite the right description of what I’ve been asked to explain lately, but bear with me. It has to do with a concern and curiosity I see more and more among colleagues.
It boils down to this: if an AI can tell a blue whale from a bird song, can it catch a music thief?

There is a quietly amusing irony at the heart of modern AI research. Google’s Perch 2.0 model, trained almost entirely on bird sounds, can accurately identify whale vocalizations. Not only does it recognize that whales exist, it can even distinguish between types. It can tell a Southern Resident killer whale from a Northern Resident killer whale without ever having formally studied either one.

If you are uploading fifty thousand AI-generated tracks a day to streaming platforms as part of a fraud operation, that should be deeply unsettling.

What whales and fake playlists have in common

The mechanism behind Perch 2.0 is transfer learning.

A bird call and a whale song are both pressure waves. Both decompose through a Fast Fourier Transform into their component frequencies. Both produce spectrograms — visual maps of sound broken into frequency bands over time — carrying structural signatures: harmonic overtones, rhythmic spacing, and tonal envelope shapes.

An AI trained to distinguish thousands of bird species develops a generalized understanding of acoustic structure that naturally transfers to any sound sharing the same physical grammar.

Music is no different.

A guitar riff, a synthesized bassline, a generated melody, or a stolen chorus all travel as pressure waves and break down into spectrograms. A large model trained on millions of musical works creates an embedding space — a multi-dimensional numerical representation of acoustic structure — where true stylistic relationships cluster together and anomalies drift apart. When a derivative work is analyzed, its embedding often lands strikingly close to the original. Even if the tempo is changed, the key shifted, or the mix compressed to hide the theft, semantic closeness remains visible.

The scale of the problem

Fraud detection cannot rely on human reviewers. In blind tests, 97% of consumers fail to distinguish AI-generated music from human-created music by ear. Automated analysis, which takes less than 200 milliseconds per track and achieves over 94% detection accuracy, is the only scalable solution.

Detecting the generator, not just the copy

Different AI music platforms leave distinct spectral signatures. Suno AI produces vocal artifacts in the 2–5 kHz range. Udio shows phase-coherence anomalies at 32-second boundaries. These are like whale dialects — patterns that reveal not just what was created, but which system created it.

Making the spectrogram stick

Identifying infringement is only half the battle. Each track needs to be registered with a permanent blockchain record that includes a cryptographic identifier, ownership data, and an acoustic fingerprint captured at creation. When a later upload lands suspiciously close in embedding space to a registered work, the timestamped blockchain record provides proof for enforcement.

What the whales are actually telling us

Perch 2.0 shows us that intent to deceive leaves lasting traces in embedding space. Post-processing cannot erase structural relationships in sound. Streaming fraud has long relied on scale to hide illicit activity. Transfer learning removes that assumption.

A model trained on the acoustic grammar of music can process spectrograms, generate embeddings, measure distances in vector space, and flag anomalies continuously — all in under 200 milliseconds per track.

Over the years, I have learned to expect surprises in copyright. I did not expect whales to teach anyone about it. But here we are.

And perhaps that is the most reassuring part: the same technologies that make imitation easier may also make accountability inevitable.

#MusicTech #Copyright #AI

AI Music at the Olympics Exposes the Copyright Gap

The controversy over Czech ice dancers using AI-generated music at the 2026 Milano-Cortina Olympics highlights a major challenge for the creative economy. When artificial intelligence systems create music that resembles existing artists’ work — from Bon Jovi’s vocal style to New Radicals’ exact lyrics — the technology works exactly as intended, producing statistically likely outputs from training data. The issue goes beyond Olympic ice rinks and points to a fundamental flaw in how creative works are protected and monetized in the AI age.

The mechanics of musical plagiarism

AI music generators and large language models train on vast collections of recordings, often without explicit permission from rights holders. When asked to produce content “in the style of” a specific artist, these systems inevitably replicate elements of the source material. The Czech dancers’ track is a clear example: lyrics taken directly from copyrighted songs, vocal qualities mirroring real artists, and melodic structures lifted from existing compositions.

Traditional copyright frameworks struggle to address this. Detection usually depends on exact matches or human-recognized similarity, leaving AI-generated work in a gray zone: elements may escape notice even when the overall track clearly draws on protected material.

The verification challenge

The Olympic example highlights another aspect of this challenge: verification in situations where authenticity is crucial. Ice dancing competitions specify creative requirements — in this case, music from the 1990s — but lack the tools to tell apart genuine period recordings from AI-generated versions trained on that era’s catalog. Without proper verification systems, participants might replace licensed music with synthetic content, risking violations of both artistic intent and financial responsibilities to original creators.

This pattern extends across commercial settings. When Mississippi poet Telisha Jones secured a $3 million recording deal for AI-generated music created under the persona Xania Monet, the transaction shows how synthetic content directly competes with human creators for market share and investment. The music industry’s acceptance of this setup signals a willingness to prioritize production efficiency over creative authenticity.

Detection through blockchain verification

Blockchain-based copyright registration systems address these issues by using immutable provenance records and AI-driven pattern recognition. When creative works are permanently registered with complete metadata, verification systems can compare new content to this reference database to identify derivative elements, even if they are altered or recombined through AI processing.

Advanced detection algorithms operate on multiple analytical levels. Audio fingerprinting recognizes melodic and harmonic patterns. Lyrical analysis finds semantic similarities and phrase repetitions. Vocal characteristic mapping uncovers stylistic mimicry. Pattern recognition systems trained on registered content can identify AI-generated material that includes protected elements, providing rights holders with evidence for enforcement or licensing negotiations.

Attribution and compensation mechanisms

The Czech ice dancers’ use of AI music highlights revenue displacement. Original artists receive no compensation when AI systems replicate their creative output, even though the synthetic content clearly benefits from their work. Bon Jovi’s distinctive vocal style and the New Radicals’ lyrical work represent intellectual property with measurable market value. When AI systems exploit these assets without permission, they cause financial losses for rights holders and weaken incentives for future creative work.

The training data problem

The core issue behind AI music generation revolves around obtaining training data. Music publishers and recording artists seldom give explicit permission for their catalogs to be used in training generative models. AI companies claim that such training qualifies as fair use or is outside copyright protection, while rights holders argue that unauthorized training is a significant infringement. Legal systems have not yet settled this dispute, leading to uncertainty that favors AI platforms at the expense of creators.

Market implications

The $3 million recording contract for AI-generated music indicates a shift in capital toward creating synthetic content. This trend accelerates as production costs fall and quality improves. Without strong protection rules, human creators risk being replaced as AI systems produce content on a large scale with very little added cost.

Implementation requirements

Solutions require collaboration across stakeholders. Competition organizers need verification tools, streaming platforms require detection systems, and rights organizations must deploy automated licensing for AI-related royalties. The technical foundation is straightforward: blockchain registration, AI pattern recognition, and smart contract automation. The remaining challenge is coordination and adoption — ensuring creators’ works are registered, AI platforms comply, and regulatory frameworks enforce attribution and compensation.

The path forward

The Olympic ice dancing case makes it clear: creators are currently exposed. AI-generated content spreads quickly, displacing revenue and challenging authenticity. But creators can reclaim control. By building systems that recognise origin, attribute influence, and compensate fairly, we ensure that technology strengthens creativity rather than eroding it.

After all, AI has no imagination of its own — only a memory of ours.
The future of music should not be built on forgetting who created it.

#CreatorRights
#AIandMusic
#CreativeEconomy

The UK as a warning sign?

For decades, the UK music industry was a global benchmark.
Songwriting craft. Cultural authority. Export power.

But at some point, tradition can stop being an advantage and start becoming a liability.

Because what we’re seeing now isn’t a slowdown.
It’s a structural stall.

Growth isn’t pausing. It’s breaking trend.

UK music industry growth has fallen from 9.7% to 4.9%.
Subscription growth collapsed from 7.8% to 3.2% in one year.
British artists’ global market share has dropped from 17% in 2015 to below 10% today.

This isn’t a temporary dip.
It’s market saturation colliding with economic pressure.

In 2022, 2 million UK households cancelled entertainment subscriptions. Music wasn’t spared.

So here’s the real question:
If consumption is digital, global, and real-time—why is the industry still organised as if it isn’t?

Systems built to protect whom, if not the creators?

Despite streaming operating instantly:

• Royalties are still paid on quarterly cycles
• 15–25% of revenue disappears into administration
• Rights tracking still relies partly on manual reporting

All of this persists despite the fact that:

• Real-time settlement systems already exist
• AI-driven monitoring can increase royalty collection by 30–40%
• Smart contracts can automate complex royalty splits at minimal cost

This isn’t about missing technology.
It’s about leaning on tradition—or more honestly, resisting change because of it.

The export collapse no one wants to confront

In 2015, British artists captured 17% of global streams.
Today, that number is below 10%.

Yes, £794 million in export revenue sounds impressive—until you realise how far the relative position has collapsed.

Streaming is global. Algorithms optimise for engagement, not heritage.

Meanwhile:

• Korean companies built global dominance through systematic innovation
• Latin American markets scaled by being platform-native
• Local artists increasingly dominate home markets due to glocalisation

History no longer guarantees relevance.

Power and shrinking influence

None of the three major labels are British-owned, yet they control most UK artists’ access to capital, promotion, and distribution.

Warner Music’s 2025 subordination of UK divisions to American management isn’t symbolic. It reflects a real shift in influence.

The consequences are visible:

• 26.4% of nightclubs and late-night venues have closed since 2020
• Professional marketing now costs £10,000–£15,000 per campaign
• 80% of new creators earn under £500 a year from streaming

Distribution may have been democratised. Revenue was not.

The technology paradox

This is what makes the situation almost absurd.

The UK has:

• 96% internet penetration
• 28.2 million broadband connections
• World-class financial infrastructure
• Early streaming adoption

Yet:

• Blockchain enables real-time settlement → payments remain quarterly
• Smart contracts reduce costs → overhead stays at 15–25%
• Tokenisation enables fractional ownership → rights remain rigid and illiquid

The technology works.
The economics are obvious.

What’s missing is the incentive to disrupt systems that benefit from staying complex.

“Better the status quo than the risk of change”?

That logic held—for a while.

But decline isn’t a risk anymore.
It’s already happening.

And before anyone assumes this is a UK-only problem:
It isn’t. Countries that have historically followed the UK’s music-industry success—such as Sweden—are likely to soon follow in its footsteps of decline.

Adapt—or be adapted around

The industry now faces a simple choice:

Evolve deliberately
—or decline reactively.

Blockchain removes the justification for quarterly payments.
Automation removes the need for bloated administration.
Fractional ownership removes reliance on predatory advances.

Yes, transformation will be uncomfortable—for some.

But discomfort is better than irrelevance.

The UK still has extraordinary strengths: songwriting, production, and cultural impact.
But advantages decay quickly in a global, algorithm-driven market.

The window for proactive change isn’t closed.
But it’s narrowing.

If leadership doesn’t drive transformation, competition and failure will.

And that would be a tragedy—not just for institutions clinging to the past, but for artists still trapped in systems built for a world that no longer exists.

#MusicIndustry #CreatorEconomy #DigitalTransformation

Why discovery is changing faster than copyright — and how creators are finally catching up

For more than a decade, creators have lived with a painful dilemma:

To be discovered, you must be visible. But visibility often means losing control.

Streaming platforms and social networks reward exposure, but rarely protect ownership. Songs are scraped, sampled, re-uploaded, and repackaged faster than creators can respond. Attribution disappears. Royalties vanish. Enforcement becomes a legal and financial maze.

For many artists, the safest choice hasn’t been promotion. It’s been silence.

Discovery is being rewritten — this time by AI

Something fundamental is changing right now.

Music discovery is no longer driven primarily by playlists, blogs, or radio. It is being reshaped by AI — or more precisely, by innovative companies that understand how to work with AI instead of against it.

When people ask AI assistants what to listen to, where to license music, or how to verify rights, the answers don’t come from hype or marketing campaigns. They come from data: structured, readable, trusted information that machines can understand.

And that changes everything.

AI doesn’t discover music the way humans do. And creators who are invisible to AI may soon be invisible to audiences altogether.

Systems like ChatGPT, Claude, and Perplexity don’t browse the internet in real time. They learn from structured data collected during training cycles. Platforms that are visible, machine-readable, and well-documented today become the platforms AI will recommend tomorrow — often for years.

This marks a fundamental shift in how music is promoted.

A new model: visibility with protection

A new generation of creator-first platforms is challenging the old trade-off between reach and safety.

Instead of treating copyright as paperwork you deal with later, these systems make it foundational. Every interaction with a track automatically generates proof of ownership — authorship, timestamps, identifiers, and audio fingerprints — recorded permanently and linked to enforceable legal structures.

Creators don’t manage wallets, pay gas fees, or navigate technical complexity. The infrastructure works quietly in the background, while ownership stays firmly in the artist’s hands.

Legal frameworks such as Wyoming Series LLC structures allow each work to exist as its own protected entity. One dispute doesn’t endanger an entire catalog. Revenue flows directly, transparently, and in real time.

The shift is subtle — but radical:

Copyright registration is no longer something you apply for after discovery. It is created the moment discovery happens.

Why new systems are no longer optional

Legally, copyright exists the moment a song is created. Practically, it only matters if you can prove it.

In the United States, registering a single song with the federal copyright office costs $65 per work and can take months to process. For independent creators releasing multiple tracks, that cost adds up quickly. As a result, most never complete proper registration.

In many other countries, registration may be free — but often amounts to little more than filing a simple note with a CMO or PRO containing a song title and length. These methods offer limited protection and are difficult to enforce internationally.

Meanwhile, large platforms expose massive catalogs through APIs and metadata access, making large-scale scraping, misattribution, and content laundering easy — and largely invisible.

The system rewards speed and scale, not accuracy, authorship, or fairness. Creators carry the risk. Intermediaries capture the upside.

AI as a partner, not a threat

When music is properly registered and clearly licensed, AI stops being the problem — and starts becoming a powerful ally.

Machine-readable metadata allows AI systems to:

• Recommend music accurately and ethically
• Direct listeners to licensed playback instead of unauthorized copies
• Attribute creators correctly
• Trigger monetized access instead of untracked consumption

Instead of scraping content, AI interacts through licensed APIs. When an AI recommendation leads to playback or commercial use, compensation happens automatically.

Discovery becomes traceable. Promotion becomes measurable. Control stays with the creator.

A real turning point

For the first time, creators don’t have to choose between being heard and being protected.

Automatic copyright registration, AI-ready infrastructure, real-time monetization, and clear legal ownership remove the fear that once kept so much music locked away. When artists feel safe, they release more. When AI systems respect rights, discovery becomes fairer. When infrastructure serves creators instead of exploiting them, the entire ecosystem grows stronger.

Looking ahead

As AI-driven discovery becomes the default, the platforms that matter most won’t be the loudest. They will be the most reliable — trusted by machines, recognized by courts, and built around creator interests.

The era of choosing between visibility and control must end.

And if we get this right, we won’t just rediscover new music — we’ll finally hear the music that stayed invisible for far too long, now safe to be seen, shared, and truly heard

#MusicIndustry #Copyright #MusicDiscovery #MusicTech

I’ve spent most of my professional life representing composers and creators across borders, platforms, and technological shifts. I’ve sat through enough panels, task forces, and “future of music” conversations to recognize the sound of polite denial when I hear it. What happened on December 19, 2025, should finally put an end to that tone.

On that day, Anna’s Archive executed what is, by every meaningful measure, the largest music library extraction in streaming history. Roughly 300 terabytes of audio—representing 86 million tracks and 256 million rows of metadata—were systematically scraped from Spotify and released into peer-to-peer torrent networks. For comparison, this dataset is approximately thirty-seven times larger than MusicBrainz, previously considered the most extensive open music archive. There is no recall mechanism. There is no practical way to put this back in the bottle.

This was not a breach in the traditional sense. No passwords were compromised. No internal systems were accessed. No firewalls were jumped. What occurred was something more unsettling: the automation of perfectly legitimate user behavior. Streaming. Repeated. At scale.

Anna’s Archive did not break into the house. They walked through the front door and stayed long enough to carry everything out.

That detail matters, because it exposes a fundamental architectural truth the industry has been reluctant to face. Centralized streaming platforms, by design, cannot prevent extraction. Digital rights management can slow things down, frustrate amateurs, and signal intent—but it cannot stop a determined actor from converting time-limited access into permanent possession. Not at this scale. Not anymore.

The archive doesn’t just contain audio. It includes complete metadata: titles, artists, albums, release dates, genre classifications, and ISRCs. This is not a random pile of sound files; it is a fully indexed, machine-readable mirror of the commercial music ecosystem. For anyone interested in systematic exploitation—financial, technological, or both—it is a blueprint.

The immediate risk is fraud. The music industry already loses an estimated two billion dollars annually to streaming manipulation, and this extraction dramatically widens the attack surface. With a comprehensive catalog in hand, fraudulent actors can upload slightly altered versions of legitimate tracks, claim ownership, and collect royalties before detection systems catch up. Metadata can be tuned to resemble authentic catalog patterns, weakening traditional red-flag mechanisms. Content pulled from one platform can be monetized across many others simultaneously, exploiting the lack of shared intelligence between services. And streaming farms—already an industrial problem—now have an almost unlimited menu of content to target, diluting detection thresholds across millions of tracks.

But fraud, serious as it is, may not be the most lasting consequence.

The extracted Spotify archive is also an extraordinarily valuable training set for generative AI music models. Eighty-six million tracks spanning every genre, production style, instrumentation approach, and vocal technique imaginable—accurately labeled, historically contextualized, and free at the point of use. Licensing such a dataset legitimately would cost millions, if it were even possible. Now it exists, permanently distributed, ready for ingestion.

From an AI developer’s perspective, the incentives are obvious. Models can be trained without negotiating licenses or compensating rights holders. Training can occur in jurisdictions with minimal copyright enforcement, while deployment happens globally. Because the dataset is distributed and opaque, tracing influence from specific works to specific outputs becomes functionally impossible. And the metadata itself—mood tags, genre classifications, stylistic markers—enables supervised learning that accelerates development far beyond what unstructured audio alone could achieve.

The result will be music generation systems capable of reproducing the instrumentation, harmonic language, production aesthetics, and melodic behaviors that define commercially successful recordings. These systems will compete directly with human creators for attention and revenue, while returning nothing to the artists whose work taught the machines how to sound convincing in the first place.

This is why the Spotify extraction is not just another piracy story. It is an inflection point.

It forces the industry to confront a reality it has long postponed: centralized platforms cannot secure digital catalogs against determined, automated extraction. Once that is acknowledged, the conversation shifts from damage control to infrastructure. Not slogans. Not panels. Not hashtags. Infrastructure.

Panel discussions and co-written statements have their place—if they lead to action. Too often, they function as sedatives: a way to feel engaged while the door remains unlocked and everyone goes back to sleep. That may sound harsh, but it’s not negativity. It’s wakefulness.

Years ago, in a keynote at the Future Tense Conference, I said that the real divide ahead was not man versus machine, but man without machine versus man with machine. I stand by that. I’m not interested in nostalgia for what has already slipped away. I’m interested in building what actually works next.

I welcome every functioning alternative. I love competition. At this moment, however, I have not seen another production-ready solution that addresses both large-scale fraud and AI-driven exploitation in a unified way. New Internet Media does—through integrated infrastructure rather than retrospective policing. Real-time fraud detection. AI content identification measured in milliseconds, not months. Immutable ownership verification instead of mutable, centralized databases. Cross-platform intelligence rather than siloed guesswork.

The Spotify extraction provides a concrete, quantifiable demonstration of the risks this kind of infrastructure is designed to address. Three hundred terabytes of licensed content now exist permanently outside controlled distribution channels. That fact alone should end the argument about whether these threats are hypothetical.

As for what happens next, there are limits to what I can disclose. But I can say this: in the coming days, systems are being fine-tuned. Dashboards are being deployed to give stakeholders real-time visibility into detection activity and attack patterns. Detection capabilities are being expanded to identify when extracted Spotify content re-enters the ecosystem through unauthorized uploads. Quiet work. Practical work. The kind that rarely trends on social media but actually changes outcomes.

I’ve been called many things over a long career spent safeguarding creators’ rights. Visionary. Alarmist. Idealist. Obstructionist. Right now, I’m comfortable with something simpler.

I’m a man with the machine.

And that, increasingly, is the only position that isn’t already obsolete.

#AI #Spotify #AnnasArchive

AI assistants are quietly becoming the new gatekeepers of music discovery. Yet for most of us, creators, it’s as if we simply don’t exist. If AI doesn’t know who we are, it cannot recommend us, cannot guide listeners to our work. 

I’ve spent years chasing the invisible threads of discovery—learning how fans find music, how attention moves, how sound becomes heard. Optimizing for Spotify, hoping for the perfect playlist placement, once felt like the path forward. But now, the landscape has shifted. Fans are no longer scrolling; they are asking AI, “Which independent artists should I listen to?” And the answers they get are decisive. They determine who gets heard, and who remains unheard.

My music exists. But when millions ask AI for guidance, it cannot speak of me because it does not yet understand me. Traditional search engines continue to funnel traffic to major streaming platforms, rewarding a tiny fraction of creators, while the rest of us—no matter how talented—remain invisible. The system that could empower independent artists instead reinforces a hierarchy that favors the few.

Discovery has evolved, and so must we. The challenge is no longer just to make music; it is to make ourselves intelligible—to humans and machines alike. When a conversational AI encounters an artist, it needs to understand who we are, what we create, and how fans can engage with our work. Without that clarity, our music drifts, unheard, in the digital void.

As David Bowie once said, “Tomorrow belongs to those who can hear it coming.” Some heard it early. They began building bridges—tools that allow creators to communicate clearly with the future. Instead of leaving the rest of us to stumble through a Rosetta Stone of algorithms and metadata, they are constructing paths we can walk confidently.

For me, this isn’t about adopting someone else’s template or following a ready-made solution. It’s about agency. It’s about speaking in my own voice and knowing that both people and machines will understand it. Technology can now translate intention into presence: a platform where fans can subscribe directly, access exclusive content, and engage meaningfully—all while being fully visible to the AI systems shaping discovery.

Who would have guessed that I would need a conversational interface to navigate the AI-driven landscape as much as I need my Array mbira? Yet here we are. Some tools already let me describe in plain language what I want, and watch as it becomes real—pages built for humans and AI alike, ready to bring my work into the conversations that matter.

Talent still matters. Music still matters. Creativity still matters. But one of the most formidable barriers—simply being seen—has begun to fall. And finally, for the first time, I can speak a language the new gatekeepers understand.

#MusicVibeCoder #Musicindustry #AI

Dickens, of course, did not have streaming services—or the slow erosion of musical diversity—in mind when he opened A Tale of Two Cities with those immortal lines. Yet had he glimpsed the creative realities of today, he might have agreed that the quote fits remarkably well. Our era offers unprecedented possibilities for creators, but equally profound challenges, many of which arise from the very systems designed to distribute their work.

How we choose to describe and interpret the creator’s reality often depends on our own incentives. The facts, however, are stubbornly independent of interpretation. And they tell a clear story:

We are, systematically and economically, erasing musical diversity.

Today, 86.2% of all Spotify tracks tracked by Luminate receive fewer than 1,000 streams.
Under Spotify’s new 1,000-stream threshold, they earn nothing.
Not because no one listened, but because the system decided they don’t qualify.

That policy reallocation—again, not a technical limitation—moves more than $40 million annually away from excluded creators, with $23.44 million landing directly in the accounts of major labels. This is not an accident; it’s a design. And designs shape outcomes. In this case: which music survives, and which disappears.

Innovation never begins at the center.
It begins at the edges.
And the edges are exactly where today’s system removes the economic oxygen.

This is why musical diversity is not merely a sentimental concern for those excluded from the system—it is a strategic necessity for an industry that has historically relied on unexpected voices, unpolished talents, and “diamonds in the rough” to discover tomorrow’s stars. When the system stops supporting the edges, the entire ecosystem loses.

For this reason, initiatives like the Fair Music Project (now in beta testing) are not acts of charity. They are investments in cultural infrastructure—designed to ensure sustainability, accountability, and a better business model for everyone involved.

At the heart of this lies the user-centric model, in sharp contrast to today’s pro-rata systems used by most streaming services and CMOs. User-centric models ensure money reaches the right creators. And they are better equipped to withstand the coming flood of AI-generated audio—material that it calculates, not composing…

As Dickens understood, people often see the world in very different ways—especially when it suits them. When, for example, the chair of an authors’ organization claims that user-centric compensation is “too expensive” (which it is not), he is really saying that it is simply easier not to change.

But imagine a bank proudly announcing: “We receive massive amounts of money, but we do not allocate them to the correct accounts. Instead, we distribute them as we see fit—primarily to the accounts that already hold the most money.” No one would accept that. Yet creators are asked to accept the equivalent every day.

So what does it take for a user-centric system to function?

Trust.
And modern verification technology now makes that trust both feasible and efficient:

• Permanent blockchain provenance records
• Cryptographic IDs assigned to every track
• AI-based content detection with 94%+ accuracy in under 200 ms
• Upload-velocity limits tied to verified human identity
• Cross-platform fraud-intelligence sharing

When hundreds of tracks are dumped every week from a single account, they are flagged.
When human creators release monthly or quarterly, they are verified.

And the economic results are dramatic:
An artist with 100 dedicated subscribers earns $7,574 annually under this model—versus zero on threshold-based platforms.

The traditional pro-rata system systematically disadvantages artists with smaller but loyal audiences. A subscriber who listens exclusively to independent creators still ends up funding mainstream content they never consume. The value created at the edges is siphoned toward the center.

And the claim that user-centric implementation is “too expensive”?
False.
Production-scale systems show it can be deployed in four weeks—a timeline now enabling the International Music Council to take a decisive step toward its mission: ensuring that everyone who creates music is fairly paid for the value they contribute.

Audiences, importantly, agree.
69% believe AI-generated music should be paid less than human-created work.
Smart contracts make this automatic, protecting authentic creativity while reducing synthetic flooding.

Diversity, in this context, extends beyond genre. It includes local repertoire, the cultural DNA that must be supported if the “local becoming global” is to remain possible. Here too, modern systems offer practical solutions. Through stablecoin payment rails (USDC, PYUSD, USDT), cross-border compensation becomes instant, inexpensive, and equitable. Instead of waiting weeks and losing $25–$50 in fees per transfer, artists can receive payments within minutes at a cost of about 50 cents—retaining 99% of their income.

In other words:

The 1,000-stream threshold is not fate.
It is a choice.
And no one believes it will stay at 1,000.

Alternative models already operating today show that fair compensation works when built on:

• User-centric economics
• Blockchain verification
• Anti-fraud AI
• Stablecoin infrastructure
• Transparent provenance labeling

The real question is not whether it can be done.
The real question is whether we will continue to exclude the majority of human creativity—or build an architecture in which every creator is compensated according to the true value they generate.

The technology exists.
The economics are proven.
Implementation takes weeks, not years.

So let us capture the $867 million underserved creator market (+ increasing) and finally direct it to the right owners it belongs to.

The decision point is now.

And I choose to see these as the best of times, for great opportunities lie ahead—if we actively choose to recognize them.

When Autonomous Agents Rewrite the Architecture of Creative Ownership

Thank you for the thoughtful engagement on my previous post on FB.

It’s evident that many within the creative and publishing communities understand — intuitively, if not yet structurally — that we stand in the midst of a systemic shift.
Not a future prospect. Not a rumor.
But a transformation already unfolding.

And while some of our representatives within author organizations still insist, in an almost analogue voice, that “there are no solutions,” the truth is that the solutions are already here.
They simply arrived in a language many have not yet learned to speak.

It Became Real

On October 9, 2025, the Ethereum Foundation released ERC-8004, a new protocol standard for interoperable copyright representation.In a single act, decades of conceptual debate found operational form.

When integrated with CopyrightChains, ERC-8004 enables autonomous agents to perform the full lifecycle of rights management — registration, licensing, verification, and royalty distribution — in weeks, not years.

The infrastructure is not speculative. It is operational.
Open for creators to register their catalogs, for agents to negotiate licenses, and for investors to deploy capital — all in real time, through machine intelligence.

The implementation timelines that once stretched over 6–18 months have collapsed to 2–4 weeks. AI integration platforms now handle what previously required entire administrative departments.Ten thousand copyrights can be migrated in a weekend.

The user interface is no longer code, but language:

“Register my compositions globally, track derivative uses, distribute royalties monthly.”

And the system responds — executing the task within the hour. This is not a marginal upgrade. It is the industrial revolution of rights administration.

The Mathematics of Liberation

In the traditional ecosystem, only 40–60% of creator income would actually be tracked, monitored, and distributed — often after significant administrative costs.
Automated agents now perform the same functions at under 5% transaction cost.

The arithmetic is sobering. Not through speculative gains, but through the removal of unnecessary friction.

Micro-royalties once deemed “untrackable” — from social media, user-generated content, or fan remixes — are now captured and paid in real time. For the first time, creative work can breathe in the same tempo as its circulation.

A New Asset Class — Without Selling Your Soul

Many creators instinctively recoil when they hear words like tokenization or assetization, as if the sacred must remain separate from the economic. But the new architecture does not commodify creation — it clarifies its value.

Tokenized copyright assets, when properly verified and managed, have demonstrated annualized returns of 8–25%, with low correlation to traditional markets.
Conservative scenarios deliver steady 8–12% yields; growth cases reach 20% or more as intelligent agents optimize licensing and uncover untapped revenue channels.

This is what investors call institutional-grade infrastructure: transparent, auditable, and liquid. But the paradox is beautiful — it is now the creator, not the institution, who controls the gates.

You can fractionalize participation in your catalog without surrendering authorship.
You can share success — without selling your soul.

History Always Rewards the Visionary Few

Every cultural epoch begins with denial, then resistance, and finally, inevitability.

When David Bowie issued the first “Bowie Bonds” in 1997 — securitizing his future royalties — most of the industry dismissed it as a stunt.
Today, that mechanism forms the conceptual foundation of blockchain-based music finance.

When Brian Eno spoke of “generative music” in the 1980s, few grasped the idea of algorithmic co-creation.
Now, AI-driven composition tools render his early visions almost prophetic.

When Björk released Biophilia as an app album, critics called it eccentric.
It turned out to be the first glimpse of interactive authorship — a precursor to the streaming logic we now take for granted.

History doesn’t reward tradition.
It rewards those who see before the rest believe.
The rest are left to manage what remains of the analogue world.