Artificial intelligence is rapidly changing the music industry.

What initially looked like a direct threat to record labels, songwriters, performers and copyright owners is increasingly becoming something else: a commercial opportunity.

The transition is revealing.

The major record companies have spent the last few years challenging AI companies over the unauthorized use of copyrighted recordings and compositions.

Yet, at the same time, some of those same companies are now negotiating — and in several cases signing — licensing agreements with AI music platforms.

The issue, therefore, is no longer simply whether the music industry is “for” or “against” artificial intelligence.

The more important question is who controls the technology, who controls the rights, and who gets paid when AI becomes part of the music business.

From Lawsuits to Licensing

The conflict became particularly visible in 2024, when Universal Music Group, Sony Music Entertainment and Warner Music Group took legal action against AI music companies including Suno and Udio.

The labels argued that copyrighted recordings had been used to train generative AI systems without authorization.

The fundamental concern was that AI could reproduce characteristics derived from existing music while creating enormous quantities of competing content.

The legal battle has not disappeared.

In August 2026, for example, independent publisher Round Hill Music filed lawsuits against Anthropic and Suno, alleging unauthorized use of copyrighted lyrics and compositions in AI training.

But alongside litigation, another path has emerged: licensing.

Universal reached an agreement with Udio in 2025, while Warner Music Group subsequently reached an agreement with Suno.

These developments point toward a new model in which copyrighted music can become licensed training or creative material rather than simply something AI companies use without permission.

That distinction is crucial.

The industry does not necessarily appear to be trying to stop AI music from existing.

It appears increasingly interested in making sure that AI music operates inside a system in which the rights holders remain part of the economic equation.

The Real Shift: AI as a Business Rather Than a Threat

This is where the story becomes particularly interesting.

A technology can be considered dangerous when it operates outside an existing business model.

The same technology can become extremely attractive once that business model is adapted to include it.

For the major labels, this means moving from:

“AI is using our music.”

to:

“How can AI use our music legally — and how can that create revenue?”

That is a very different question.

Universal Music Group has pursued collaborations around AI-powered music creation, while the major labels have explored licensing arrangements with AI companies.

At the same time, Spotify and Universal have moved toward licensed AI-powered fan experiences involving covers and remixes.

The commercial logic is relatively straightforward.

If AI-generated music is going to become a significant part of the market, the companies that already control enormous music catalogues have a strong incentive to participate in that market rather than simply watch it develop from the outside.

The Numbers Show Why the Industry Cannot Ignore AI

The scale of AI-generated music is already enormous.

In June 2026, Deezer reported that AI-generated tracks accounted for more than 50% of all new daily uploads on the platform, reaching approximately 90,000 AI-generated tracks per day at peak levels.

That does not mean that half of all music being listened to is AI-generated.

It means something different — and potentially more important.

The cost and speed of music production are changing dramatically.

A human artist may spend weeks or months developing a project.

Generative AI can produce thousands of musical variations in a fraction of that time.

This creates an entirely new economic proposition: more music, produced faster, at dramatically lower marginal cost.

And whenever the cost of producing content falls, the entertainment industry inevitably begins asking how that content can be packaged, distributed and monetized.

The Rise of the Virtual Artist

Perhaps the most striking development is not AI-generated songs themselves, but the emergence of AI-driven musical identities.

One of the most discussed examples is Xania Monet, an AI-created R&B artist developed by Telisha “Nikki” Jones.

The project attracted considerable attention after entering the US music ecosystem and eventually securing a record deal reportedly worth up to $3 million with Hallwood Media.

The significance of Xania Monet goes beyond one artist.

It demonstrates that an AI-generated musical character can potentially become a commercial property.

The character can have:

• a name;
• a visual identity;
• a voice;
• a musical style;
• a catalogue;
• social-media content;
• a narrative;
• and, potentially, an entire commercial ecosystem.

At that point, the industry is no longer simply selling songs.

It may be selling a scalable digital artist brand.

An Artist Who Exists as a Digital Property

This possibility creates a fascinating new economic model.

A traditional artist is a human being.

Their availability is limited.

They need time to record, perform, travel and rest.

Their career is also shaped by personal circumstances.

A virtual artist operates differently.

A digital character can theoretically release music continuously, appear in multiple forms of media, evolve visually and musically, and exist simultaneously across music, video games, advertising, social media and merchandise.

This does not automatically make a virtual artist better.

Nor does it mean audiences will abandon human performers.

But it changes the economics of experimentation.

A label can potentially test musical concepts, identities and audiences at a speed that would be difficult — and expensive — to reproduce using traditional artist development.

And this is precisely why AI-generated artists deserve attention.

Not because they are necessarily replacing human musicians today, but because they could become a new category of intellectual property.

The Uncomfortable Paradox

This is where the apparent contradiction becomes difficult to ignore.

The music industry argues that artists should be protected when their voices, recordings, compositions and identities are used without permission.

That principle is understandable.

But the same industry is increasingly exploring how those very assets can participate in licensed AI systems.

There is nothing inherently contradictory about that from a legal or commercial perspective.

The distinction is authorization.

Using an artist’s work without permission and using it under a negotiated licence are fundamentally different situations.

The real question is therefore not whether AI should exist.

It is:

Who gives permission, under what conditions, and who receives the money?

From Protecting the Catalogue to Monetizing the Catalogue

For decades, the music industry’s economic power has been built around intellectual property.

Master recordings, publishing rights, artist names, brands and catalogues represent enormous long-term assets.

Generative AI introduces a new possibility: those assets can potentially become inputs into entirely new forms of music creation.

That creates a potentially powerful business model.

Instead of seeing a catalogue only as something people listen to, the industry can potentially see it as something that can also be licensed, transformed, remixed and used to generate new experiences.

This is why licensing agreements are so important.

They could transform AI from a technology that threatens the value of existing catalogues into another mechanism for extracting value from them.

In simple terms:

the catalogue does not disappear. Its commercial function expands.

But There Is Another Side to the Equation

The rapid growth of AI-generated music also creates a serious problem: abundance.

When music becomes extremely cheap to produce, the market can become saturated.

Deezer’s data illustrates the scale of the phenomenon.

The platform has developed systems to detect AI-generated music and has taken steps to prevent AI-generated tracks involved in streaming fraud from generating royalties.

This distinction matters.

AI itself is not necessarily the problem.

The problem is industrial-scale production designed primarily to exploit recommendation systems, streaming economics or royalty mechanisms.

If someone can generate 10,000 songs and automatically upload them, the economics of streaming can potentially be manipulated.

The result could be an internet filled with music that nobody particularly wanted — but which exists because producing it became almost free.

The Real Competition May Be for Attention

This could ultimately become more important than the competition between AI and human musicians.

Music has always been abundant.

What has never been abundant is human attention.

If AI allows the industry to produce exponentially more songs, the scarce resource will not be music.

It will be discovery.

Which song gets recommended?

Which artist receives editorial support?

Which virtual character becomes visible?

Which human artist gets a playlist placement?

Which release receives advertising?

Which track appears in a film, game or social-media campaign?

In an environment where millions of songs can be generated, control over discovery may become even more valuable than control over production.

What Happens to the Human Artist?

This is perhaps the most important question.

A professional human artist carries costs that an AI-generated project may not.

There are studio sessions, musicians, producers, engineers, photographers, videos, promotion, management, touring and years of work required to develop an identity.

AI can reduce some of those costs dramatically.

That does not mean human artists will disappear.

It does mean that the economic threshold for creating a commercially viable music project could become much lower.

The industry may therefore enter a period in which thousands of new musical identities can be created and tested.

Some will fail.

Some will become viral.

And a few may become extremely valuable.

The selection process could become increasingly data-driven.

The Question of Authenticity

There is also a cultural question that technology cannot answer.

Does the audience care whether the person behind a song actually exists?

The answer appears to be: sometimes yes, sometimes no.

People have always connected with fictional characters, anonymous producers, masked performers and carefully constructed stage identities.

A virtual artist is therefore not necessarily an impossible concept.

But AI introduces another layer.

If the voice, performance and visual identity are generated by machines, what exactly is the audience connecting with?

The song?

The character?

The human creator behind the character?

Or simply the story surrounding it?

The answer may vary from artist to artist.

Transparency Could Become the New Currency

As AI music becomes increasingly difficult to distinguish from human-created music, transparency may become commercially important.

Listeners may eventually want to know whether a track was:

100% human-created, AI-assisted, or substantially AI-generated.

The distinction becomes even more complicated as AI is increasingly used inside conventional production.

A producer might use AI for a vocal idea, a synthesizer part, drums, arrangement, mastering or sound design while the final track remains a fundamentally human production.

Recent research is already exploring methods for estimating the proportion of AI-generated material within hybrid recordings rather than treating music as simply “AI” or “human.”

That may become increasingly relevant as the technology matures.

The Industry Is Not Choosing Between AI and Humans

Perhaps the most misleading way to describe the current situation is as a battle between artificial intelligence and human creativity.

The reality is more complicated.

The music industry is likely to contain all of these models simultaneously:

• human artists using AI as a tool;
• human artists working without AI;
• AI-generated music created under licence;
• virtual artists managed by humans;
• and mass-produced AI content created primarily for scale.

The commercial battle will be over which of these models audiences actually value.

A New Music Industry Is Taking Shape

The most interesting development is therefore not that major labels have suddenly “changed their minds” about artificial intelligence.

They haven’t.

Their position is becoming more sophisticated.

They are challenging unauthorized use of their intellectual property while simultaneously exploring ways to participate in the AI economy.

That is not necessarily hypocrisy.

It is what large commercial organizations tend to do when confronted with a technology capable of changing the economics of their industry.

They try to control the risk.

Then they try to control the opportunity.

And finally, if possible, they try to monetize both.

The Question Is No Longer Whether AI Will Enter Music

AI has already entered music.

The numbers make that clear.

With tens of thousands of AI-generated tracks appearing every day, licensing agreements between major rights holders and AI companies, and virtual artists already attracting serious commercial investment, the experiment has moved well beyond the laboratory.

The next stage will be much more consequential.

Who owns the AI artist?

Who owns the voice?

Who owns the training data?

Who gets paid?

Who controls the catalogue?

And, perhaps most importantly:

Who decides what the audience gets to hear?

The future of music may not be a world without human artists.

It may be a world in which human artists, AI tools and virtual performers coexist — while the companies controlling the infrastructure compete to determine how that new ecosystem is monetized.

The irony is that the technology once feared for its ability to disrupt the music business may ultimately become one of the music industry’s most valuable commercial tools.

The battle, in the end, may not be about stopping AI. It may be about owning the future that AI creates.

Major labels once challenged generative AI. Now they are learning how to license it, monetize it, and build new artists around it.

Sources:

• Reuters — Major labels, AI music lawsuits and licensing
• Reuters — Round Hill Music lawsuits against Suno and Anthropic
• Deezer — AI music exceeds 50% of daily uploads
• Forbes — Xania Monet and the $3 million record deal
• Music Business Worldwide — Suno, Udio and major-label licensing agreements
• The Wall Street Journal — Major labels negotiate AI licensing deals
• Reuters — German court ruling involving Suno and copyright
• Research paper — Quantifying AI-generated material in hybrid music

From legal battles over copyright to licensed AI platforms and virtual artists — how the music industry is turning artificial intelligence from a potential threat into a new commercial opportunity.

AI may change how music is created. The real question is who controls the technology, who owns the rights, and who benefits from the new economy it creates.

From the protection of traditional catalogues to the rise of virtual artists, the boundary between human creativity, technology and commercial opportunity is rapidly changing.