Intelligence Piracy: Why Copyright’s Next Battle Is Not About Copying

By Oliver Walsh, Head of Evidence Services, Rouse

Someone stole the musician.

For more than a century, copyright law has focused on protecting creative works.

A song, a film, a photograph, a book. Digital piracy simply changed the method of copying. CDs became MP3s. DVDs became torrents. Streaming platforms replaced peer-to-peer networks. The infringement evolved, but the underlying act remained the same. Someone copied the work.

Artificial intelligence changes something far more fundamental.

Today, an AI model can study an entire catalogue, absorb its musical language, learn its creative decisions and generate new works that sound unmistakably familiar without reproducing any individual recording.

Nothing has been copied in the traditional sense, yet something valuable has clearly been taken. The question is no longer whether a particular song was illegally reproduced. The question is whether the creative intelligence behind thousands of songs has been extracted.

That is a fundamentally different problem. I believe it deserves a different name.

I call it "Intelligence Piracy".

Copyright has always protected outputs.

AI, however, attacks capability, and I believe that distinction matters.

When an artist spends thirty years building a catalogue, they are not simply producing thousands of songs. They are creating a body of knowledge:

>Patterns.

>Relationships.

>Stylistic decisions.

>Creative judgement.

A modern language model can consume that entire body of work and distil it into statistical representations capable of producing remarkably similar outputs.

No single song needs to be replicated, as the ability itself has been reestablished. The possible theft, if it occurred, no longer concerns the completed product. Instead, it is about the knowledge necessary to create the next one.

Every catalogue is becoming a teacher model.

This is where many discussions around AI and copyright become trapped. The debate often centres on whether specific outputs infringe copyright. Whilst that question remains important, I believe it is also becoming the wrong starting point, as every substantial creative archive now serves another purpose.

A music catalogue is no longer simply a collection of recordings.

= It is training data.

A television archive is not merely entertainment.

= It is behavioural information.

A library of illustrations is not just artwork.

= It is an instruction.

Every catalogue has become a teacher model, and that shift changes the economics of infringement.

Traditional piracy competed with legitimate distribution. Intelligence piracy competes with future creation.

Three generations of digital piracy.

Digital piracy has evolved in distinct stages.

The first generation targeted content.

>Songs.

>Films.

>Books.

>Software.

The objective was simple: copy and distribute.

The second generation targeted platforms.

>Illegal IPTV services.

>Pirate streaming operations.

>Subscription resellers.

The content became secondary. The infrastructure became the commercial asset.

Artificial intelligence introduces a third generation.

The objective is no longer to steal the content or the platform. It is to acquire the capability that produced the content in the first place. The asset under attack has changed. That changes how investigators, lawyers and rights holders must think about infringement.

The music industry is simply first.

Music happens to be the first major creative sector confronting Intelligence Piracy. It will not be the last.

Film libraries.

Sports data.

News archives.

Scientific research.

Product designs.

Software repositories.

Legal databases.

Every sector that has spent decades building valuable knowledge assets is discovering that those assets can teach machines.

We need a new language.

For thirty years, digital piracy meant making unauthorised copies. Artificial intelligence forces us to consider a different possibility...

Perhaps the most valuable thing being taken is not the work itself. Perhaps it is the accumulated intelligence that made the work possible. If that is true, then copyright is entering a new phase.

Not because copying has disappeared. But because copying is no longer the whole story. The next generation of IP disputes will increasingly revolve around capability, attribution and evidence.

The challenge is not simply protecting creativity. It is protecting the intelligence behind it.

The "Intelligence Piracy" Series.

This article is the first in a series exploring Intelligence Piracy and what it means for rights holders, investigators and policymakers. In the coming articles, we will examine how these questions are already emerging across film and television, sport, news and publishing, brands, and the AI industry itself. Each sector tells a different story, but they all point towards the same conclusion: the next generation of IP disputes will be shaped not only by what was copied, but by what was learned.